A
The Astrophysical Journal, 659:1265 Y 1290, 2007 April 20
# 2007. The American Astronomical Society. All rights reserved. Printed in U.S.A.
s-PROCESS ABUNDANCES IN PLANETARY NEBULAE
Brian Sharpee,1 Yong Zhang,2, 3, 4 Robert Williams,2 Eric Pellegrini,5 Kenneth Cavagnolo,5
Jack A. Baldwin,5 Mark Phillips,6 and Xiao-Wei Liu 3
Received 2005 May 6; accepted 2006 December 4
ABSTRACT
The s-process should occur in all but the lower mass progenitor stars of planetary nebulae, and this should be reflected in the chemical composition of the gas that is expelled to create the current planetary nebula shell. Weak forbidden emission lines are expected from several s-process elements in these shells and have been searched for and in
some cases detected in previous investigations. Here we extend these studies by combining very high signal-to-noise
ratio echelle spectra of a sample of PNe with a critical analysis of the identification of the emission lines of Z > 30 ions.
Emission lines of Br, Kr, Xe, Rb, Ba, and Pb are detected with a reasonable degree of certainty in at least some of the
objects studied here, and we also tentatively identify lines from Te and I, each in one object. The strengths of these
lines indicate enhancement of s-process elements in the central star progenitors, and we determine the abundances of
Br, Kr, and Xe, elements for which atomic data relevant for abundance determination have recently become available.
As representative elements of the ‘‘light’’ and ‘‘heavy’’s-process peaks, Kr and Xe exhibit similar enhancements over
solar values, suggesting that PN progenitors experience substantial neutron exposure.
Subject headingg
s: ISM: abundances — nuclear reactions, nucleosynthesis, abundances —
planetary nebulae: general
Online material: machine-readable table, tar file
1. INTRODUCTION
IR fine-structure nebular emission from postYFe peak ions
( Dinerstein 2001; Sterling & Dinerstein 2004) and from far-UV
resonance-line absorption by s-process elements in the intervening nebular shell seen in Far Ultraviolet Spectroscopic Explorer
(FUSE ) spectra of their central stars (Sterling el al. 2002; Sterling
& Dinerstein 2003). They detected IR emission from Se and Kr in
roughly 50% of their sample PNe from which abundances of up
to 10 times solar were derived for these elements. The FUSE
spectra revealed Ge iii absorption in five PNe for which Ge enhancements of 3Y10 times solar were deduced, clear evidence
that central star AGB progenitors are major sites of s-process
element production.
As part of our ongoing program to detect and identify the
weakest lines in PNe at visible wavelengths down to levels significantly below that of the continuum, viz., <105 the intensity
of H, we have obtained very high signal-to-noise ratio (S/ N )
spectra at high spectral resolution of some of the higher surface
brightness nebulae. Some fraction of the weaker lines observed
are likely to originate from postYFe peak elements; therefore, we
have used the automatic line identification routine EMILI (Sharpee
et al. 2003), supplemented by recent energy level data for heavier
elements, to assist in the search for such lines. Since the publication of PB94, new calculations have been made for spontaneous emission coefficients (Biémont et al. 1995) and collision
strengths (Schöning 1997; Schöning & Butler 1998) for fourthand fifth-row elemental ion transitions. In this study we use these
newer atomic parameters to compute ionic and overall elemental
abundances for Br, Kr, and Xe in order to make comparisons
with their abundances in H ii regions and the Sun and to derive
s-process enrichment factors relevant to the study of s-process
nucleosynthesis in the progenitor stars.
As remnants of stars that have evolved through the asymptotic
giant branch (AGB) phase, most planetary nebulae ( PNe) are believed to consist of material that has undergone nuclear processing in the precursor star via the s-process. The analysis of nebular
emission from elements that have experienced nucleosynthesis in
the parent star provides valuable information for stellar models.
However, the detection of emission lines from ions enhanced
by the s-process has been hampered by their weakness and by
uncertainties in the atomic data needed for the analysis, including line wavelengths.
An initial attack on this problem was made a decade ago by
Péquignot & Baluteau (1994, hereafter PB94), who obtained a
deep optical spectrum of the high-ionization PN NGC 7027. Using the best atomic data available for the energy levels of the
more prominent ionization stages of elements in the fourth and
fifth rows of the periodic table, they identified a number of postYFe
peak emission lines in the nebula. From the observed line intensities they concluded that the elements Kr and Xe, the latter
normally a predominantly r-process element in stars of solar metallicity, were enhanced in NGC 7027 by factors of 20 relative to
their initial formation abundances, presumably roughly solar.
They also detected Ba ii and [Br iii] emission at intensities indicating that Ba could be enhanced whereas Br might be depleted
relative to solar values, subject to uncertainty due to poorly known
excitation cross sections.
Dinerstein and collaborators have subsequently pursued the
study of s-process abundances in PNe through surveys to detect
1
SRI International, Menlo Park, CA 94025.
Space Telescope Science Institute, Baltimore, MD 21218.
Department of Astronomy, Peking University, Beijing 100871, China.
4
Current address: Department of Physics, University of Hong Kong, Hong
Kong, China.
5
Department of Physics and Astronomy, Michigan State University, East
Lansing, MI 48824.
6
Las Campanas Observatory, Carnegie Observatories, Casilla 601, La Serena,
Chile.
2
3
2. OBSERVATIONS AND DATA REDUCTION
Our present sample of objects consists of four PNe ( IC 2501,
IC 4191, NGC 2440, and NGC 7027) that satisfy the most important criteria for the detection of weak emission lines in having
(1) high surface brightnesses, (2) low expansion velocities, and
1265
1266
SHARPEE ET AL.
Vol. 659
TABLE 1
Journal of Observations
Object
NGC 7027........................
IC 2501 ............................
IC 2501 ............................
IC 4191 ............................
IC 4191 ............................
NGC 2440........................
Dates
( UT )
Telescope
KPNO 4
LCO 6.5
LCO 6.5
LCO 6.5
LCO 6.5
LCO 6.5
m
m
m
m
m
m
2002
2003
2003
2003
2003
2003
Jun 19Y22
Feb 12Y13
Feb 25Y26
Feb 12Y13
Feb 25Y26
Feb 12Y13
(3) except for NGC 7027, relatively small internal dust extinction. The latter two criteria produce sharper lines and higher peak
intensities relative to the continuum.
We obtained spectra of the PNe IC 2501, IC 4191, and
NGC 2440 during two observing runs of two nights each in
early 2003 using the Las Campanas Observatory ( LCO) Baade
6.5 m telescope with the MIKE echelle spectrograph. Similar
instrumentation setups were used during the two runs. MIKE is a
dual-beam spectrograph that simultaneously measures separate
red and blue spectra, giving useful data over the continuous
wavelength ranges 3280Y 4700 and 4590Y7580 8 at resolutions
(k/k) of 28,000 and 22,000, respectively, for the 100 slit width
we used.
A series of long and short exposures were taken of each nebula
on one or more nights, typically adding up to a period of 2Y3 hr
at a time spent observing each object. Because MIKE is used
without an image rotator at a Nasmyth focus, the orientation of
the slit with respect to each nebula rotated by a large amount during these series of exposures. The central star was placed along a
line perpendicular to the spectrograph slit so that the slit center
was roughly midway between the central star and outer edges of
the nebula, and then the telescope was guided to keep the star in
that same position as seen on the acquisition/guide TV. The result
was that during the course of the observations the slit swept out
an arc about the central star, so that the final spectrum integrates
over an area of the nebula that is much larger than the 1 00 ; 5 00 slit.
The total integration times in the combined long exposures were
630 minutes for IC 2501, 450 minutes for IC 4191, and 330 minutes for NGC 2440. A journal of the observations is given in
Table 1 and includes the approximate position of the center of
the slit with respect to each of the central stars at the start of the
series of exposures.
The optics of the MIKE spectrograph introduce strong distortion into the image formed on the detector, so the projected emission lines have significantly different tilts as a function of their
position on it. When extracting spectra of objects such as these
PNe that fill the slit, the tilts introduce unacceptable smearing in
the wavelength direction (typically 3 pixels over the 40 pixel slit
length) unless they are corrected for during the extraction of the
one-dimensional spectra from the two-dimensional echelle image. We wrote our own set of auxiliary FORTRAN programs to
calibrate the tilts and perform the proper extraction. This procedure was tested on the emission lines in the comparison lamp
spectrum. Compared to the emission line from a single pixel at
the slit center, lines summed over the full slit length came out at
the same pixel location in the dispersion direction and were broadened by 2% on average. We are thus confident that the effects of
these tilts are negligible in the extracted PN spectra. These tiltcorrected spectra were then fed into the same suite of IRAF-based
reduction programs used with the NGC 7027 data as described
next.
Integration Times
(s)
Slit Center
34 ; 1200, 9 ; 300, 1 ; 120, 4 ; 60, 6 ; 30
6 ; 1800, 1 ; 300, 4 ; 60
15 ; 1800, 2 ; 60
5 ; 1800, 1 ; 1200, 1 ; 300, 1 ; 60
10 ; 1800, 1 ; 300, 3 ; 60
11 ; 1800, 1 ; 300, 1 ; 60
3.500 W, 0.500 N
200 E
200 E
200 E
200 E
1400 E
NGC 7027 was observed on four nights in 2002 June with
the Mayall 4 m Telescope at Kitt Peak National Observatory
(KPNO) using the Cassegrain echelle spectrograph. Because we
were interested in detecting faint 40Y50 km s1 FWHM emission lines over the widest possible wavelength range, rather than
detailed measurements of the line profiles, we used the short
UV camera with the 79.1 groove mm1 echelle grating, crossdispersing grating 226-1 in first order, and a GG-475 order separating filter. This gave full wavelength coverage from 4600 to
9200 8 at 20 km s1 FWHM (k/k ¼ 15;000) resolution
with our 200 slit width, with partial coverage out to 9900 8. On
each night we offset to the same position with the slit at P:A: ¼
145 and centered 0.500 north and 3.500 west of the central star,
which is the brightest part of the PN shell in H emission. Most
of the observing time was spent obtaining sequences of 1200 s
exposures that added up to a total of 680 minutes of integration
time. We also took a number of shorter exposures 30, 60, 120,
and 300 s in length to measure bright emission lines that were
saturated on the longer exposures.
The data for all four PNe were reduced using standard procedures with IRAF-based reduction packages in the same manner
that has been described in detail in our discussion of comparable
spectra of the PN IC 418 that we obtained previously with the
Cerro Tololo Inter-American Observatory 4 m echelle spectrograph (Sharpee et al. 2004). A correction was made for the presence of Rowland ghosts near strong lines (Baldwin et al. 2000).
To correct for flexure and temperature drift effects in the spectrographs, comparison lamp spectra were taken at roughly 1 hr
intervals, and the wavelength calibration for each PN spectrum
was made using the comparison lamp taken nearest to it in time.
The wavelength fits indicate a 1 uncertainty in the wavelength
scale that varies over each echelle order, and it is of the order
4Y6 km s1 for all of our spectra. The spectra were flux-calibrated
using observations of several spectrophotometric standard stars
from Hamuy et al. (1994).
The final extracted flux- and wavelength-calibrated spectra of
all the PNe were used for our line identification and analysis and
are available in FITS format as an electronic supplement to the
present article.
3. EMISSION-LINE SELECTION
The line selection procedure is initiated by fitting the continuum of each echelle order of the final spectrum with a smooth
function that is pegged to the observed continuum at approximately 10 wavelength intervals distributed over each order. We
developed an automated procedure that selects what should be
reliable continuum points for the fitting by selecting wavelength
regions with a paucity of lines. Because bad pixels, artifacts caused
by scattered light in the spectrograph due to strong emission lines,
and line blends can cause poor fits, the fits were reviewed manually for each order and adjusted as necessary to ensure that the
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1267
TABLE 2
Nebular Parameters
Fig. 1.— Portion of the IC 2501 spectrum illustrating the automated continuum fitting procedure. The vertical lines designate emission features defined by
the ED fitting algorithm.
continuum representation was valid. An example of the output
of the automatic continuum fitting algorithm is shown in Figure 1,
where a typical fit prior to manual adjustment is displayed. Once
the proper continuum level is established, each order is sampled
pixel by pixel to detect emission features, which are defined as
regions where the observed flux exceeds the continuum flux by
7 or more over a wavelength interval equal to or greater than
that of the resolution of the spectrograph.
All putative emission features were examined individually by
eye and compared with their appearance on the original twodimensional echelle images to establish their reality. We have
found that all features with fluxes greater than 12 of that of the
continuum, i.e., with S/N > 12, are clearly visible on the echelle
images. Since scattered light features usually trail across multiple orders, one can generally distinguish such artifacts from real
lines by visually examining the images. The most uncertain aspect of the line detection process is distinguishing multiple line
blends in real emission features. The expansion velocities of PNe
are such that intrinsic line widths do vary by factors of 2Y3 with
the level of ionization, so it can be difficult to discriminate between one broad line and two or more closely spaced narrow
lines. Figure 1 gives an example of the features within a selected
wavelength region in one of our PN spectra that have been designated by our software as emission lines.
For detection of the weakest nebular lines, where the distinction between a continuum noise spike and a real feature can be
difficult to establish with certainty, we have intercompared the
spectra of those PNe that have similar levels of ionization but
different radial velocities. Noise spikes do not generally produce
features having the width of the instrumental resolution, and instrument artifacts tend to occur on the same place on the detector,
which is at different rest wavelengths in the PN spectra because
of their differing radial velocities. Since spectra were obtained on
different instruments (MIKE and the 4 m KPNO echelle), all
significant scattered light ghosts could be detected through comparisons between the PN spectra obtained with MIKE and the
NGC 7027 spectrum. Our final line lists do contain a few weak
features having S/N < 7 that are present in at least two of the
objects.
The telluric nightglow emission spectrum was also sampled
by these spectra, as sky subtraction in these extended objects was
deemed impractical, particularly in regard to the subtraction adding significant additional noise. However, most nightglow lines
were distinguishable on the two-dimensional spectra by their uniform intensity and characteristic shape and size, namely, those of
the imaging slit. The positions and intensities of prospective nightglow lines were compared to those listed in the telluric feature
Parameter
NGC 2440
IC 2501
IC 4191
NGC 7027
V( geo) ( km s1) ........
c( H ).........................
+66.7
0.55
+23.2
0.55
40.1
0.77
7.1
1.17
atlases of Osterbrock et al. (1996) and Hanuschik (2003) and
likely matches removed from the nebular list except in those
cases where their blending with stronger nebular lines was
deemed only a minor contaminant. For lines considered as candidate s-process ion transitions, nightglow sources were given
additional scrutiny through a comparison with the original spectra of the Hanuschik (2003) atlas, the likely identifications of the
atlas lines (Cosby et al. 2006), time-averaged observed intensities of their constituent systems (Cosby & Slanger 2007), and in
some cases model spectra of those systems constructed with the
molecular simulation code DIATOM.7
Following the selection of sets of likely nebular lines in each
of the PNe, their wavelengths and intensities were determined by
single or multiple (in the case of blended features) Gaussian fitting of their profiles and immediate underlying continua. Simple
summing was used for the most irregular and ill-defined line profiles with intensity-weighted centroids utilized as wavelengths.
Emission lines appearing in multiple orders were then collated
and their attributes averaged together. Emission-line intensities
and wavelengths from short- and long-duration exposure spectra
of each PN were then normalized to a particular fiducial frame.
Line wavelengths were shifted to the nebular rest frame through
a comparison with either the Balmer (MIKE PNe) or Paschen
( NGC 7027) sequence laboratory wavelengths. Line intensities
were dereddened with the Galactic extinction law of Howarth
(1983) utilizing an iterative process involving the magnitude of
either the Balmer or Paschen jump, as well as the Balmer or He ii
5Yn sequence decrements to establish electron temperatures and
c (H ) logarithmic extinction at H values. The c(H ) values
for each PN are in good agreement with those listed in Cahn et al.
(1992). Errors in the final line intensities in all PN spectra, as
determined from the formal errors to the profile and continua
fits, are similar to those found for NGC 7027: 41% for I <
105 I( H ), 20% for I ¼ 105 I( H ) to 104I( H ), 11% for
I ¼ 104 to 103, and <6% for I > 103 I(H ) on average. This
is independent of any errors arising from the reddening correction. For weak lines, where the greatest contributor to uncertainty
is the indeterminate level of the true continuum, the formal errors
in the fit probably understate the actual uncertainty in the intensity. From random inspection of several lines at the 105 H intensity level, it is likely that some lines may have uncertainties
ranging upward to 100% in regions where the continuum level is
rapidly undulating, complicated by scattered light artifacts from
adjacent orders, or affected by strong telluric absorption. Final
geocentric offsets and c( H ) values are presented in Table 2.
Figure 2 presents a histogram showing the fraction of lines in
each of the nebulae observed with MIKE for different intensity
levels relative to H that we have determined to be real and that
also appear in at least one of the other two PNe (the NGC 7027
spectrum is omitted here due to its different wavelength coverage
and spectral resolution). The strongest emission lines are all detected in each of the nebulae. Even at intensities down to 105 H,
where S/N 10 typically, roughly 50% of the weak features
identified in the individual PNe also appear in one of the other
7
See http://www-mpl.sri.com /software/ DIATOM / DIATOM.html.
1268
SHARPEE ET AL.
Fig. 2.— Fraction of lines present in the spectrum of each PN that are also detected in another of the PNe. The numbers give the total number of detected lines
in each flux interval.
nebulae, suggesting that they are true nebular lines. The present
spectra are among the deepest emission spectra taken of nebulae,
revealing some of the weakest lines yet observed. This can be
seen in Figure 3, where the cumulative number of lines exceeding a given flux level relative to H is shown for several of
the most extensive PN spectral studies published in the recent
literature.
4. PLASMA DIAGNOSTICS AND ABUNDANCES
Electron temperatures and densities are presented in Table 3.
The IRAF NEBULAR package task temden (Shaw & Dufour
1995) was used to equate relative intensities of collisionally
excited diagnostic lines to densities and temperatures by matching a diagnostic for each from ions of similar ionization energy
and solving for self-consistent values. Each ion was modeled
by a five or greater level atom for this purpose, with spontaneous emission coefficients and collision strengths for the five lowest levels drawn primarily from the compilation of Mendoza
(1983). Although these are not the default values currently distributed with NEBULAR, this atomic data set yielded good
agreement in temperature and density among diagnostics in
our previous analysis of IC 418 and was utilized in the original
NEBULAR release. Spontaneous emission coefficients from
Froese Fischer & Tachiev (2004) and collision strengths from
Wilson & Bell (2002) were used for N i and Cl ii, respectively.
For the remaining energy levels and for the ions Cl iv and K v,
the atomic data used in the most recent release of NEBULAR
were utilized. The departure of the [ K v] density diagnostic from
other density diagnostic values, particularly in NGC 2440 and
IC 4191, may indicate errors in these atomic data. However,
s-process elemental abundances derived later using this diagnostic ( Kr+4/ H+ and Xe+5/ H+) were completely insensitive to
density. Errors in diagnostic values were determined by selecting the extrema values from computations at all combinations
of diagnostic line ratios plus and minus their (1 ) uncertainties,
including an error estimate for the reddening correction. Indeterminate error limits, such as occurred when a ratio value
exceeded the asymptotic limit or where the paired diagnostics
failed to converge at a particular ratio value, are listed without a
value. Balmer (MIKE PNe) and Paschen (NGC 7027) jump temperatures were calculated in the manner described by Zhang et al.
(2004), and as reported by them for many PNe, they are lower
Vol. 659
Fig. 3.— Cumulative number of observed lines exceeding a given flux level
for recently published nebular spectra. We consider only those lines within the
wavelength range 3510Y7470 8, which is covered by all the spectra.
than those temperatures determined from the collisionally excited
lines.
Ionic abundances derived from both collisionally excited and
nominal radiative recombination lines are presented in Table 4.
The IRAF NEBULAR package tasks ionic and abundance
were used to make computations from the collisionally excited
lines listed in Table 3 using temperatures and densities derived
from the diagnostics with the closest ionization potential to each
ion. Where no diagnostics where clearly appropriate, averaged
values for temperature and density were used. Uncertainties were
computed in the same manner as the diagnostic values, by calculation of the abundance at every combination of temperature,
density, and line intensity ratio plus or minus their (1 ) uncertainties (where available) and selecting the extrema values. Uncertainties were not calculated for abundance computed with
averaged diagnostics values. For the recombination lines, effective recombination coefficients were combined with line intensities to make abundance determinations, in the manner described
by Sharpee et al. (2004) with temperature and densities again
drawn from diagnostics nearest in ionization potential to each
ion. For He+/ H+, the coefficients of Smits (1996) or Benjamin
et al. (1999) were used to determine an average abundance from
the k4923, k5876, and k6678 lines, with corrections for collisional excitation (case A for triplets, case B for singlets) taken
from Kingdon & Ferland (1995), while coefficients from Storey
& Hummer (1995) were used to calculate average He+2/H + abundances from various He ii lines. To determine C +2, N +2, O+2, and
Ne+2 abundances relative to H +, recent effective recombination
coefficients (Storey 1994; Kisielius et al. 1998; Davey et al. 2000;
Kisielius & Storey 2002) for the strongest observable multiplets
were used to calculate abundances. As seen in Table 4, O+2/H +
and Ne+2/H + abundances deduced from the collisionally excited
lines are systematically lower than those derived from radiative
recombination lines, as is generally the case in PNe ( Liu 2006;
Robertson-Tessi & Garnett 2005).
5. EMISSION-LINE IDENTIFICATION
The emission-line identification code EMILI (Sharpee et al.
2003) was used to make the majority of emission-line identifications. EMILI creates models of the ionization energyYdependent
velocity field and ionization level of a PN or H ii region from
user-supplied empirical data. These models are used by EMILI
to select from a large atomic transition database all transitions
within 5 times an observed line’s wavelength measurement error
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1269
TABLE 3
Electron Temperatures and Densities
Diagnostic
NGC 2440
IC 2501
IC 4191
NGC 7027
::
21000:15000
11000
11000þ9000
4000
11000
8500þ2100
1600
9300þ2100
1600
7400þ2200
1500
9900þ2200
1700
...
::
12000:7000
...
9000þ6000
3000
...
12000þ3000
2000
11800þ2300
1900
11000þ2400
1800
13000þ3000
2000
41000þ17000
11000
::
15000:8000
...
...
...
47400þ1900
1800
...
49200þ1300
1200
...
...
7900þ500
300
...
...
12500 1100
13000þ800
600
9900 300
8600þ500
400
11600þ300
200
11600þ1400
900
8000
11300þ300
200
...
...
...
12900 200
...
13700 200
...
...
8000
Density (cm3)
[ N i] k5198/k5200.........................................
[S ii] k6716/k6731 .........................................
[O ii] k3726/k3729a .......................................
[O ii] k3726/k3729b .......................................
[Cl iii] k5517/k5537.......................................
[Ar iv] k4711/k4741c .....................................
[Ar iv] k4711/k4741d .....................................
[Ar iv] k4711/k4741e .....................................
[ K v] k4123/k4163........................................
3300þ2700
1300
3100þ1200
700
3300þ1100
800
3700
4700þ1100
900
6200þ1600
1300
5900þ1600
1300
6300þ1600
1300
43000þ19000
13000
Temperature ( K)
[O i] (k6300+k6464)/k5577...........................
[S ii] (k6716+k6731)/(k4068+k4076) ...........
[O ii] (k3726+k3729)/(k7320+k7330)...........
[ N ii] (k6548+k6583)/k5755 .........................
[Ar iii] (k7136+k7751)/k5192 .......................
[O iii] (k4959+k5007)/k4363 ........................
[Cl iv] (k7531+k8046)/k5323 .......................
[ Ne iii] (k3869+k3968)/k3343 ......................
[Ar v] (k6435+k7005)/k4625 ........................
Balmer/ Paschen ( NGC 7027) jump .............
a
b
c
d
e
9400 300
15000þ6000
4000
16000
þ700
12600600
13100þ700
600
14700þ600
500
12900þ1000
700
15500þ500
400
16400þ1500
1100
11000
Vs. [O ii] temperature diagnostic.
Vs. [ N ii] temperature diagnostic.
Vs. [O iii] temperature diagnostic.
Vs. [Cl iv] temperature diagnostic.
Vs. [ Ne iii] temperature diagnostic.
TABLE 4
Ionic Abundances
X+i/ H+
NGC 2440
IC 2501
IC 4191
NGC 7027
Collisionally Excited Lines
N0/ H+ ............
N+/ H+ ............
O0/ H+ ............
O+/ H+ ............
O+2/ H+ ..........
Ne+2/ H+ .........
S+/ H+.............
S+2/ H+ ...........
Cl+/ H+ ...........
Cl+2/ H+..........
Cl+3/ H+..........
Ar+2/ H+ .........
Ar+3/ H+ .........
Ar+4/ H+ .........
K+4/ H+ ..........
7:25þ0:21
0:16
7:85þ0:05
0:06
7:76þ0:06
0:05
7.24
8:26þ0:04
0:05
7:41þ0:03
0:04
5.5 0.3
6:23þ0:07
0:08
...
4.70 0.06
4:69þ0:06
0:07
6.13 0.04
5.70 0.03
5:24þ0:06
0:08
4:59þ0:09
0:11
::
6:9:0:5
þ0:16
7:020:10
::
7:68:0:14
7.30
þ0:04
8:590:05
7.58 0.03
5.93
þ0:11
6:580:12
3.43
þ0:09
4:920:10
þ0:15
4:660:21
6.27 0.07
4.75 0.03
...
...
::
6:4:0:4
þ0:12
6:740:09
::
7:34:0:11
7:5þ0:3
0:2
8.79 0.05
þ0:03
7:790:04
5.67
þ0:07
6:080:09
3.18
þ0:07
4:560:08
þ0:06
5:230:07
þ0:05
5:700:06
þ0:03
5:740:04
þ0:09
4:570:12
þ0:14
3:770:17
::
6:1:0:3
7.19
::
7:28:0:04
7.68
8.44
...
5.80
6.32 0.02
3.94
4.76 0.02
4.72 0.02
6.09 0.02
5.77 0.02
5.60
...
11.00
10.08
8.46
8.28
9.08
8.76
10.81
10.60
8.84
7.89
8.46
...
Recombination Lines
He+/ H+ ..........
He+2/ H+ .........
C+2/ H+ ...........
N+2/H+ ...........
O+2/ H+ ..........
Ne+2/ H+ .........
6900þ300
200
12000
13000
10800þ900
1100
9400þ600
500
9500þ300
200
6100þ900
500
11000 200
...
7000
10.86
10.72
8.65
8.43
8.51
8.08
11.04
10.61
8.91
8.04
8.69
8.26
Note.— In units of 12 þ log ( X/H).
and to compute relative intensities for emission lines corresponding to those transitions. Those transitions predicted to produce
emission-line intensities within 3 orders of magnitude of the highest value among all transitions initially selected are then subjected
to a test for the presence of lines from the same LS-coupled multiplets (but not all lines from the same upper level) at expected wavelengths and relative intensities. Potential identifications are then
assigned a numeric likelihood parameter based on their wavelength agreement with the observed line, strength of the predicted
emission line, and results of the multiplet check, and they are
ranked and presented.
Many emission lines from Z > 30 elemental ions were observed in NGC 7027 by PB94. To place these lines on an equal
footing for identification purposes with those arising from more
abundant lighter elements, EMILI used the Atomic Line List version 2.05 of P. van Hoof 8 as its reference database, which extends
to Z ¼ 36 (Kr). The latest experimental determinations available
in the literature for ground electron configuration energy levels
of Z > 36 ions were then added to this database. Transition wavelengths for the mostly optically forbidden transitions among these
levels were constructed by differencing the level energies. Sources
for all Z > 30 ion energy levels and associated atomic data used
in subsequent analysis are provided in Table 5. All ions with
certain or probable line identifications in NGC 7027, as rated
by PB94, most of those with more tenuous identifications, and a
handful of other ions they suggested to be worthy of future consideration at a higher spectral resolution were incorporated into
the database. However, ions with level uncertainties perceived
or explicitly stated to be greater than 1.0 cm1 in their source
literature were excluded.
8
See http://www.pa.uky.edu /~ peter/newpage /.
1270
SHARPEE ET AL.
TABLE 5
Atomic Data for Z > 30 Ions
Ion
Levels
Transition Probabilities
Br iii ...........................
Br iv ...........................
Kr iii ...........................
Kr iv ...........................
Kr v ............................
Rb iv...........................
Rb v............................
Sr ii.............................
Sr iv............................
Sr v.............................
Sr vi............................
Y v .............................
Y vi ............................
Zr vii...........................
Te iii ...........................
I iii ..............................
I v ...............................
Xe iii...........................
Xe iv...........................
Xe v............................
Xe vi...........................
Cs v ............................
Cs vi ...........................
Ba ii............................
Ba iv...........................
Ba v............................
Ba vii ..........................
Ba viii .........................
Pb ii ............................
ALL 2.05a, Biémont & Hansen (1986a) (2, 3)
ALL 2.05
ALL 2.05
ALL 2.05
ALL 2.05
Persson & Wahlstrom (1985)
Persson & Petterson (1984)
Moore (1952)
Hansen & Persson (1976)
Hansen & Persson (1974)
Persson & Petterson (1984)
Reader & Epstein (1972)
Persson & Reader (1986)
Reader & Acquista (1976)
Joshi et al. (1992)
Tauheed & Joshi (1993a)
Kaufman et al. (1988)
Persson et al. (1988)
Tauheed et al. (1993)
Gallard et al. (1999)
Churilov & Joshi (2000)
Tauheed & Joshi (1993b)
Tauheed et al. (1991)
Karlsson & Litzén (1999)
Sansonetti et al. (1993)
Reader (1983)
Tauheed & Joshi (1992)
Churilov et al. (2002)
Moore (1958)
Biémont & Hansen (1986a)
Biémont & Hansen (1986a)
Biémont & Hansen (1986b)
Biémont & Hansen (1986a)
Biémont & Hansen (1986a)
Biémont & Hansen (1986b)
Biémont & Hansen (1986a)
Barge et al. (1998) (5Y1, 5Y2)
Biémont et al. (1988)
Biémont & Hansen (1986b)
Biémont & Hansen (1986a)
Biémont et al. (1988)
Biémont & Hansen (1986b)
Biémont & Hansen (1986b)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Klose et al. (2002)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Biémont et al. (1995)
Safronova et al. (2005)
a
Collision Strengths
Schöning
Schöning
Schöning
Schöning
Schöning
Schöning
Schöning
Schöning
...
...
(1997)
(1997)
(1997)
...
...
...
...
...
...
...
...
...
...
...
...
& Butler
& Butler
...
& Butler
...
...
& Butler
& Butler
...
...
...
...
(1998)
(1998)
(1998)
(1998)
(1998)
Atomic Line List version 2.05.
EMILI was run against each nebula’s set of observed emissionline wavelengths and intensities. Electron temperature and density values were provided to EMILI by the results of standard
plasma diagnostics, derived from strong collisionally excited
lines with certain identifications. Construction of the empirical
kinetic and ionization models were also drawn from those same
lines. The EMILI reference elemental abundances were set to
the solar values of Lodders (2003). EMILI was also run against
the IC 418 line list of Sharpee et al. (2003) to determine if some
of its remaining unidentified lines could be identified with an expanded set of transitions and to also act as a foil for the mostly
higher excitation PNe considered in the present sample. The line
TABLE 6
Emission-Line Identifications and Intensities
kcorr
obs
(8)
FWHM
( km s1)
F(H ¼ 100)
I(H ¼ 100)
S/ N
Line Identification
Notesa
7.2
13.9
10.4
8.4
50.0
13.3
37.0
11.5
13.6
7.1
7.1
10.7
32.2
627.6
7.1
[ Mn iii] k3315.01
Ne ii k3323.73
Ne ii k3327.15
N ii k3328.72
Ne ii k3334.84
N ii k3340.87
[ Ne iii] k3342.54
[Cl iii] k3353.21
N i k7507.61
...
He i k7513.33
C iii k7514.30
O iii k7715.99
[Cl iv] k7530.80
[ Xe iv] k7535.44
:?
IC 2501
3315.23.......................
3323.75.......................
3327.17.......................
3328.69.......................
3334.83.......................
3340.81.......................
3342.64.......................
3353.22.......................
7507.48.......................
7509.97.......................
7512.69.......................
7514.36.......................
7515.77.......................
7530.42.......................
7534.70.......................
23.73
19.16
27.98
19.13
18.57
13.75
45.56
33.40
22.85
57.84
52.14
36.15
22.81
29.59
30.21
0.0088
0.0116
0.0087
0.0056
0.0529
0.0188
0.0527
0.0071
0.0048
0.0033
0.0044
0.0038
0.0173
0.5710
0.0022
0.0200
0.0260
0.0190
0.0130
0.1200
0.0420
0.1200
0.0160
0.0027
0.0019
0.0025
0.0022
0.0099
0.3300
0.0013
:
Notes.—Table 6 is published in its entirety in the electronic edition of the Astrophysical Journal. A portion is shown here for guidance
regarding its form and content.
a
: = uncertain identification; ? = uncertain feature; bl = line blend; ns = blend with night-sky line.
TABLE 7
Possible Krypton Line Identifications
NGC 2440
Transition
[Kr iii] 4p4 3P2 Y4p4 1D2 k6826.70............................................
C i 3p 1P1 Y4d 1D2o ( V21) k6828.12.....................................
He i 3s 3S1 Y16p 3P o k6827.88..............................................
[ Fe iv] 3d 5 4P1/2 Y3d 5 2D53/2 k6826.50 ................................
Fe ii] c4F7/2 Yy2G9o/2 k6826.79...............................................
SKY: OH 7Y2 R1(3.5) k6827.46...........................................
SKY: OH 7Y2 R1(4.5) k6828.47...........................................
[Kr iv] 4p3 4S3o/2 Y4p3 2D5o/2 k5346.02 .......................................
S ii 4s0 2D3/2 Y4p0 2F5o/2 ( V38) k5345.71...............................
C iii 2s4d 3D2 Y2p3d 3P1o ( V13.01) k5345.88 ......................
[ Fe ii] a4F3/2 Yb4P3/2 k5347.65.............................................
Mg i 4s 3S1 Y9p 3P o k5345.98...............................................
Fe ii f4D7/2 Y4f 2½3o7/2 k5345.95 ............................................
Si ii 6p 2P3o/2 Y4p0 2P3/2 k5346.25..........................................
[Kr iv] 4p3 4S3o/2 Y4p3 2D3o/2 k5867.74 .......................................
Al ii 4d 3DY6f 3Fo ( V41) k5867.64, k5867.78, k5867.89 .......
Si ii 4s0 4P3o/2 Y4p0 4P1/2 ( V48) k5867.42 ..............................
Ni ii 4f 2½4o7/2 Y7d 4H9/2 k5867.99 ........................................
He ii 5Y29 k5869.02 .............................................................
[Cr iii] 3d 4 5D3 Y3d 4 3H5 k5866.97.......................................
Ni ii 4f 2½5o11/2 Y7d 2G9/2 k5867.68 .......................................
[Kr iv] 4p3 2D3o/2 Y4p3 2P3o/2 k6107.8.........................................
[ Fe ii] a6D9/2 Ya2G7/2 k6107.28 ............................................
Ca i 4s7d 3D2 Y3d20p 3P1o k6107.84.....................................
Cr ii c4F5/2 Y4p 4F1o/2 k6107.96 .............................................
[Kr iv] 4p3 2D5o/2 Y4p3 2P3o/2 k6798.4.........................................
C ii 3s0 4P3o/2 Y3p0 4D1/2 ( V14) k6798.10 ..............................
Ca i 3d5s 1D2 Y3d13p 1P1o k6798.28.....................................
[Kr v] 4p2 3P1 Y4p2 1D2 k6256.06.............................................
C ii 4p 2P1o/2 Y5d 2D3/2 ( V10.03) k6257.18...........................
C ii 3d0 2D3o/2 Y 4p0 2P1/2 ( V38.03) k6256.52 .......................
S ii 5p 4S3o/2 Y7s 4P5/2 k6256.35.............................................
Fe i a3H4 Yz3G4o ( V169) k6256.36 .......................................
O i 3p0 3F4 Y4d0 3G3o ( V50) k6256.47...................................
C ii 4p0 4D5/2 Y5d0 4D7o/2 k6256.24.........................................
Ca i 4s9s 1S0 Y3d25p 1P1o k6256.45 ......................................
[Co ii] a3D2 Ya1F3 k6256.46.................................................
SKY: OH 9Y3 Q2(0.5) k6256.94 ..........................................
SKY: OH 9Y3 Q1(1.5) k6257.96 ..........................................
[Kr v] 4p2 3P2 Y4p2 1D2 k8243.39.............................................
N i 3s 4P5/2 Y3p 4P3o/2 ( V2) k8242.39 ...................................
H i 3Y43 k8243.69 ................................................................
O iii 5g G 2[9/2]o Y6h H 2[11/2] k8244.10............................
IC 2501
V b
V
ko, I/IHa ( km s1) Multc IDId ko, I/IH ( km s1) Mult IDI
V
ko, I/IH ( km s1) Mult IDI
6826.74
1.0(5)
...
...
...
...
...
5346.04
3.5(4)
...
...
...
...
...
5867.68
4.6(4)
...
...
...
...
...
6107.68?
5.6(6)
...
...
6798.01?
2.8(5)
1.3(5)y
6256.31
4.4(5)
1.5(5)y
...
...
...
...
...
...
...
...
OUT
...
...
...
6826.39?
1.5(5)
...
...
...
...
...
5346.06
2.4(4)
...
...
...
...
...
5867.83
1.9(4)
...
...
...
...
...
6107.50
1.9(5)
...
...
...
[8.1(6)]
...
6256.31
1.5(5)
...
...
...
...
...
...
...
...
...
OUT
...
...
...
13.6
76.0
65.5
4.8
17.6
...
...
2.2
19.6
10.1
89.2
4.5
6.2
10.7
4.6
2.6
21.0
8.2
60.8
44.0
7.7
14.7
10.8
16.7
22.6
...
...
...
12.0
41.7
10.1
1.9
2.4
7.7
3.4
6.7
7.2
...
...
...
...
...
...
Ref.
Notes
1.8
60.6
50.1
10.5
2.2
...
9.2
1.1
18.5
9.0
90.3
3.4
5.1
11.8
3.1
2.0
16.4
15.8
68.5
36.3
0.0
5.9
19.6
7.9
13.8
17.2
4.0
11.9
12.0
41.7
10.1
1.9
2.4
7.7
3.4
6.7
7.2
...
12.4
...
...
...
...
0/0
...
0/0
4/0
4/0
...
...
1/1
2/0
5/0
8/0
0/0
0/0
3/0
1/1
...
6/0
...
...
5/0
...
...
3/0
...
...
2/0
7/2
0/0
0/0
2/0
2/0
1/0
5/0
4/0
4/0
0/0
1/0
...
...
...
...
...
...
5A
...
...
5A
6C
...
. . .
3A
9
7
...
4B
4B
7
3A
...
9
...
...
7B
...
. . .:
6A:
...
...
8bl
1Abl
6
5Dbl
8
5D
4A
4A
4A
5D
5D
5D
...
. . .bl
...
...
...
...
6826.82
4.6(5)
2.5(5)y
...
...
...
...
5346.07
6.1(5)
...
...
...
...
...
5867.82
8.5(5)
...
...
...
...
...
...
[4.3(6)]
...
...
6798.15
1.4(5)
...
6256.57
5.2(5)
8.0(6)y
...
...
...
...
...
...
...
...
OUT
...
...
...
NGC 7027
Transition
[Kr iii] 4p4 3P2 Y4p4 1D2 k6826.70............................................
C i 3p 1P1 Y4d 1D2o ( V21) k6828.12.....................................
He i 3s 3S1 Y16p 3P o k6827.88..............................................
[ Fe iv] 3d 5 4P1/2 Y3d 5 2D53/2 k6826.50 ................................
Fe ii] c4F7/2 Yy2G9o/2 k6826.79...............................................
SKY: OH 7Y2 R1(3.5) k6827.46...........................................
SKY: OH 7Y2 R1(4.5) k6828.47...........................................
[Kr iv] 4p3 4S3o/2 Y4p3 2D5o/2 k5346.02 .......................................
S ii 4s0 2D3/2 Y4p0 2F5o/2 ( V38) k5345.71...............................
C iii 2s4d 3D2 Y2p3d 3P1o ( V13.01) k5345.88 ......................
[ Fe ii] a4F3/2 Yb4P3/2 k5347.65.............................................
Mg i 4s 3S1 Y9p 3P o k5345.98...............................................
Fe ii f4D7/2 Y4f 2½3o7/2 k5345.95 ............................................
Si ii 6p 2P3o/2 Y4p0 2P3/2 k5346.25..........................................
IC 4191
5.3
57.1
46.6
14.1
14.1
4.8
...
2.8
20.2
10.7
88.7
5.1
6.7
10.1
4.1
2.0
20.5
8.7
61.4
43.5
7.2
...
...
...
...
11.0
2.2
5.7
24.5
29.2
2.4
10.5
10.1
4.8
15.8
5.8
5.3
5.3
...
...
...
...
...
0/0
...
0/0
4/0
4/0
...
...
1/1
2/0
5/0
8/1
0/0
0/0
3/1
1/1
...
6/0
...
0/0
5/1
...
...
...
...
...
2/0
7/5
...
0/0
2/0
2/1
1/0
5/0
4/0
4/0
...
1/0
...
...
...
...
...
...
5Abl
...
...
5A
6C
. . .bl
...
3A
9
6D
...
4B
5C
6D
4A
...
...
...
...
6B
...
...
...
...
...
7D
1A
...
7bl
8
2Abl
6
5D
5D
7
...
4B
. . .bl
...
...
...
...
...
IC 418
V
ko, I/IH ( km s1) Mult
IDI
V
ko, I/IH ( km s1) Mult IDI
6826.91
5.7(4)
4.6(4)y
...
...
...
...
5345.99
1.9(3)
...
...
...
...
...
. . .bl
...
. . .bl
7A
7A
. . .bl
...
1A
9
6
...
2B
3C
8
6826.87
3.3(4)
3.2(4)y
...
...
...
...
5345.94
3.5(5)
...
...
...
...
...
9.2
53.2
42.6
18.0
5.3
12.7
...
1.7
15.7
6.2
93.2
0.6
2.2
14.6
...
...
0/0
4/0
4/0
...
...
1/1
2/0
5/0
...
0/0
0/0
3/0
7.5
54.9
44.4
16.3
3.5
...
1.8
4.5
12.9
3.4
96.0
2.2
0.6
17.4
0/0 5Abl
. . . . . . B95, Z05
... ...
B95
4/0 7C
4/0 6B
. . . . . .bl
... ...
1/1 3A
2/0
9 H95, P04
5/0
5
Z05
... ...
ZL02
0/0
5
0/0 3A
3/0
8
e,f
f
0/0
...
0/0
4/1
4/0
...
...
1/1
2/0
5/0
8/0
0/0
0/0
3/2
1/1
...
6/0
...
0/0
5/2
...
...
3/0
...
...
...
...
...
0/0
2/0
2/0
1/0
5/0
4/0
4/0
0/0
1/0
...
...
...
...
...
...
6B:
...
...
3A
9D
...
...
3A
9
6
...
4B
5D
4B
4A
...
>9
...
...
4A
...
...
4A:
...
...
...
...
...
5D
8
5D
4A
4A
5D
4A
5D
5D
...
...
...
...
...
...
1272
SHARPEE ET AL.
TABLE 7—Continued
NGC 7027
Transition
[ Kr iv] 4p3 4S3o/2 Y4p3 2D3o/2 k5867.74.....................................
Al ii 4d 3DY6f 3F o ( V41) k5867.64, k5867.78, k5867.89.....
Si ii 4s0 4P3o/2 Y4p0 4P1/2 ( V48) k5867.42 ............................
Ni ii 4f 2½4o7/2 Y7d 4H9/2 k5867.99 ......................................
He ii 5Y29 k5869.02 ...........................................................
[Cr iii] 3d 4 5D3 Y3d 4 3H5 k5866.97.....................................
Ni ii 4f 2½5o11/2 Y7d 2G9/2 k5867.68 .....................................
[ Kr iv] 4p3 2D3o/2 Y4p3 2P3o/2 k6107.8 ......................................
[ Fe ii] a6D9/2 Ya2G7/2 k6107.28 ..........................................
Ca i 4s7d 3D2 Y3d20p 3P1o k6107.84...................................
Cr ii c4F5/2 Y4p 4F1o/2 k6107.96 ...........................................
[ Kr iv] 4p3 2D5o/2 Y4p3 2P3o/2 k6798.4 ......................................
C ii 3s0 4P3o/2 Y3p0 4D1/2 ( V14) k6798.10 ............................
Ca i 3d5s 1D2 Y3d13p 1P1o k6798.28...................................
[ Kr v] 4p2 3P1 Y4p2 1D2 k6256.06 ..........................................
C ii 4p 2P1o/2 Y5d 2D3/2 ( V10.03) k6257.18.........................
C ii 3d 0 2D3o/2 Y 4p0 2P1/2 ( V38.03) k6256.52.....................
S ii 5p 4S3o/2 Y7s 4P5/2 k6256.35...........................................
Fe i a3H4 Yz3Go4 ( V169) k6256.36 .....................................
O i 3p0 3F4 Y4d 0 3Go3 ( V50) k6256.47 ................................
C ii 4p0 4D5/2 Y5d 0 4D7o/2 k6256.24 ......................................
Ca i 4s9s 1S0 Y3d25p 1P1o k6256.45 ....................................
[Co ii] a3D2 Ya1F3 k6256.46...............................................
SKY: OH 9Y3 Q2(0.5) k6256.94 ........................................
SKY: OH 9Y3 Q1(1.5) k6257.96 ........................................
[ Kr v] 4p2 3P2 Y4p2 1D2 k8243.39 ..........................................
N i 3s 4P5/2 Y3p 4P3o/2 ( V2) k8242.39 .................................
H i 3Y43 k8243.69 ..............................................................
O iii 5g G 2[9/2]o Y6h H 2[11/2] k8244.10..........................
IC 418
V
ko, I/IH ( km s1) Mult IDI
5867.71
2.6(3)
...
...
...
...
...
6107.83
7.2(5)
...
...
6798.36
4.1(5)
1.6(5)y
6256.38
1.7(4)
1.2(4)y
...
...
...
...
...
...
...
...
8244.33?
1.4(4)
...
...
1.5
3.6
14.8
14.3
67.0
37.8
1.5
1.5
27.0
0.5
6.4
1.8
11.5
3.5
15.3
38.4
6.7
1.4
1.0
4.3
6.7
3.3
3.8
34.0
...
34.2
70.9
23.3
8.4
V
ko, I/IH ( km s1)
1/1 1A 5867.73
5/0 5C
3.5(5)
6/0
8
...
0/0
7
...
... ...
...
5/0 . . .
...
0/0 4B
...
3/1 5A
...
3/0 . . . [1.4(5)]
1/0 5A
...
9/0 5A
...
3/1 4Abl
...
7/1 7bl [2.1(5)]
0/0 5B
...
1/0 7bl
...
. . . . . .bl [8.4(5)]
2/0 6bl
...
1/0 5D
...
5/0 4A
...
4/0 5D
...
4/0
7
...
0/0 4A
...
1/0 4A
...
. . . . . .bl
...
... ...
...
. . . . . .: 8243.70
. . . . . . 4.2(4)
0/0 5A
...
0/0 7B
...
0.5
4.6
13.3
13.3
66.0
38.9
2.6
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
11.3
47.7
0.4
14.6
Mult
IDI
1/1
4/0
6/0
0/0
...
5/0
0/0
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
0/0
...
2A
6
8
7
...
...
3B
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
2A
...
Ref.
B00, S03
H01
E04
Z05
Notes
g
B95
S03
Z05
h
B95
S03
B95
a
Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ‘‘OUT’’ means not in observed range; (3) ‘‘?’’ denotes an uncertain feature. Intensity: (1) numbers in
parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features.
b
Observed (ko ) transition wavelength ( km s1).
c
EMILI multiplet check statistics: number expected /number observed.
d
EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ‘‘bl’’ ( blend). Definition of IDI given in x 6.
e
Unidentified line in IC 418 ( Baldwin et al. 2000).
f
Unidentified line in Orion Nebula (Sharpee et al. 2003).
g
NGC 2440, IC 4191, NGC 7027: identified as a separate line.
h
NGC 7027: affected strongly by telluric absorption.
References.—( B95) PN NGC 7027 ( Baluteau et al. 1995); ( B00) Orion Nebula ( Baldwin et al. 2000); ( E04) Orion Nebula ( Esteban et al. 2004); ( H95) PN
NGC 6886 ( Hyung et al. 1995); ( H01) PN IC 5217 ( Hyung et al. 2001); ( P04) PN NGC 5315 ( Peimbert et al. 2004); (S03) PN IC 418 (Sharpee et al. 2003); ( Z05)
PN NGC 7027 ( Zhang et al. 2005); ( ZL02) PN Mz 3 ( Zhang & Liu 2002).
identification lists produced by EMILI were then visually inspected, and the EMILI-preferred assignments were compared
to identifications in the literature for those same lines in PNe
and H ii region spectra of similar depth, spectral resolution, and
level of ionization. The entries in our final line list are in many
cases taken from among the highest ranked EMILI-suggested
identifications.
A segment of a final line list is given in Table 6; the full line list
is available in the electronic version of this manuscript. This
table lists all observed lines with their characteristics, viz., observed wavelength in the nebula rest frame, identifications,
FWHMs, S/N, observed and reddening-corrected intensities relative to H, and their most likely identifications. Lines with uncertain identifications are denoted by a colon, likely blends are
marked with a ‘‘bl’’ or ‘‘ns’’ if blended with a telluric emission
feature, and uncertain features are marked with a question mark.
Only the perceived strongest component of a blend is listed in the
table. Lines without obvious identifications are listed here without an identification.
6. Z > 30 LINE IDENTIFICATIONS
PB94 detected 25 emission lines from several Z > 30 ions in
their optical spectra of NGC 7027, including Se, Br, Kr, Rb, Sr,
and Y from the fourth row of the periodic table, Xe and Ba from
the fifth row, and Pb from the sixth. Eighteen of these detections
were considered ‘‘certain’’ or ‘‘probable’’ and seven considered
‘‘possible.’’ Lines from 13 additional Z > 30 ions were also either tentatively identified or proposed as future targets for spectra
with greater spectral resolution and better S/N.
Given the depth and high resolution of the spectra considered
here, confirmation of the PB94 identifications in multiple PNe was
sought, as were additional lines belonging to other Z > 30 ions.
The use of EMILI allows prospective Z > 30 transitions and
weaker transitions of more abundant lighter elements to be
treated equally for purposes of emission-line identification.
Collision strength and spontaneous emission coefficient calculations for these transitions allow accurate predictions to be
made of the relative intensities of lines arising from the same
TABLE 8
Possible Xenon Line Identifications
NGC 2440
IC 2501
b
Transition
4 3
4 1
[ Xe iii] 5p P2 Y5p D2 k5846.77..............................
He ii 5Y31 k5846.66 ................................................
[ Fe ii] a2G7/2 Yc2G9/2 k5847.32 ...............................
Fe ii 4d e4F5/2 Y4f 2½3o5/2 k5846.78..........................
Fe i x5P1o Y6d 2[3/2]1 k5846.60 ...............................
[ Xe iv] 5p3 4S3o/2 Y5p3 2D5o/2 k5709.20 .........................
N ii 3s 3P2o Y3p 3D2 ( V3) k5710.77 .........................
[ Fe i] a5D3 Ya5P1 k5708.97.....................................
Si i 4p 3D2 Y18s (3/2, 1/2)o1 k5708.91 ......................
Fe ii] y4P3o/2 Ye6D1/2 k5709.04..................................
[ Xe iv] 5p3 4S3o/2 Y5p3 2D3o/2 k7535.4 ...........................
Fe ii] b2F5/2 Yz4F5o/2 ( V87) k7534.82.......................
N ii 5f.G 2[7/2]4 Y10d 1F3o k7535.10 .......................
Ne i 3p 2[1/2]1 Y3d 2[1/2]o1 ( V8.01) k7535.77 ........
[Cr ii] b4D5/2 Yc4D5/2 k7534.80................................
Ne ii 3d0 2P3/2 Y6f 2½3o5/2 k7534.75 ..........................
[ Xe v] 5p2 3P0 Y5p2 3P2 k7076.8 .................................
C i 3p 3D2 Y4d 3D2o ( V26.01) k7076.48 ..................
[ Fe iii] 3d 6 3P41 Y3d 6 1S40 k7078.10 ......................
[ Ni ii] 4s 4F3/2 Y4s 4P3/2 k7078.04 ...........................
Ni ii 5d 4F7/2 Y4f 2½3o7/2 k7076.90............................
Ca i 4s15s 3S1 Y3d22p 3P1o k7077.02 .......................
Ca i 4s14s 1S0 Y3d17f 1P1o k7077.08........................
3 SKY: 3Y2 O2 b 1þ
g YX g P12(11) k7078.58 ......
[ Xe vi] 5p 2P1o/2 Y5p 2P3o/2 k6408.9 ..............................
He ii 5Y15 k6406.38 ................................................
[ Fe iii] 3d 6 1S40 Y3d 6 3P22 k6408.50 ......................
C iv 9Y17 k6408.70 .................................................
V
ko , I/IHa ( km s1)
5846.73
3.2(4)
...
...
...
5708.95?
1.2(5)
...
...
...
7534.70
1.3(5)
...
...
...
...
7077.02
1.5(5)
...
...
...
...
...
...
6408.69?
2.8(5)
...
...
Multc
IDId
ko , I/IH
...
0/0
3/0
0/0
...
0/0
...
7/0
0/0
7/1
0/0
0/0
0/0
0/0
4/0
0/0
...
...
0/0
...
...
...
...
...
...
...
1/0
...
...
5B
7C
3A
...
5
...
4C
3A
3A
7
2A
6
7
5
3B
...
...
5A
...
...
...
...
. . .
. . .:
...
4A
...
5846.64?
7.7(6)
...
...
...
5708.97?
6.5(6)
...
...
...
7534.71
1.6(5)
...
...
...
...
...
[5.5(5)]
...
...
...
...
...
...
...
[2.4(6)]
...
...
2.1
3.6
30.3
2.6
6.7
13.1
95.6
1.1
2.1
4.7
27.9
4.8
15.9
42.6
4.0
2.0
9.3
22.9
45.8
20.3
28.0
22.9
20.8
0.4
9.8
108.1
8.9
0.5
NGC 7027
Transition
4 3
4 1
[ Xe iii] 5p P2 Y5p D2 k5846.77..............................
He ii 5Y31 k5846.66 ................................................
[ Fe ii] a2G7/2 Yc2G9/2 k5847.32 ...............................
Fe ii 4d e4F5/2 Y4f 2½3o5/2 k5846.78..........................
Fe i x5P1o Y6d 2[3/2]1 k5846.60 ...............................
[ Xe iv] 5p3 4S3o/2 Y5p3 2D5o/2 k5709.20 .........................
N ii 3s 3P2o Y3p 3D2 ( V3) k5710.77 .........................
[ Fe i] a5D3 Ya5P1 k5708.97.........................................
Si i 4p 3D2 Y18s (3/2, 1/2)o1 k5708.91 ......................
Fe ii] y4P3o/2 Ye6D1/2 k5709.04..................................
[ Xe iv] 5p3 4S3o/2 Y5p3 2D3o/2 k7535.4 ...........................
Fe ii] b2F5/2 Yz4F5o/2 ( V87) k7534.82.......................
N ii 5f.G 2[7/2]4 Y10d 1F3o k7535.10 .......................
Ne i 3p 2[1/2]1 Y3d 2[1/2]o1 ( V8.01) k7535.77 ........
[Cr ii] b4D5/2 Yc4D5/2 k7534.80................................
Ne ii 3d0 2P3/2 Y6f 2½3o5/2 k7534.75 ..........................
[ Xe v] 5p2 3P0 Y5p2 3P2 k7076.8 .................................
C i 3p 3D2 Y4d 3D2o ( V26.01) k7076.48 ..................
[ Fe iii] 3d 6 3P41 Y3d 6 1S40 k7078.10 ......................
[ Ni ii] 4s 4F3/2 Y4s 4P3/2 k7078.04 ...........................
Ni ii 5d 4F7/2 Y4f 2½3o7/2 k7076.90............................
Ca i 4s15s 3S1 Y3d22p 3P1o k7077.02 .......................
Ca i 4s14s 1S0 Y3d17f 1P1o k7077.08........................
ko , I/IH
5846.65
4.2(4)
1.4(4)y
...
...
5708.89
1.1(4)
...
...
...
7534.94
1.7(4)
...
...
...
...
7077.00
2.0(5)
...
...
...
...
...
V
( km s1)
6.2
0.5
34.4
6.7
2.6
16.2
98.8
4.2
1.1
7.9
18.3
4.8
6.4
33.0
5.6
7.6
8.5
22.0
...
44.1
4.2
0.8
3.0
Mult
IC 4191
V
( km s1) Mult IDI
6.7
1.0
34.9
7.2
2.1
12.1
94.6
0.0
3.2
3.7
27.5
4.4
15.5
42.2
3.6
1.6
...
...
...
...
...
...
...
...
...
...
...
...
...
0/0
3/0
0/0
...
1/0
...
7/1
0/0
7/0
...
1/0
0/0
...
8/0
0/0
...
...
...
...
...
...
...
...
...
...
...
...
ko , I/IH
. . .: 5846.77
3A 5.2(5)
7C
...
4B
...
...
...
5:
5708.98?
. . . 1.2(5)
3A:
...
3A
...
5
...
. . . 7534.73
4A 3.6(5)
6C
...
...
...
6C
...
5B
...
. . . 7076.94?
. . . 3.9(6)
...
...
...
...
...
...
...
...
...
...
...
...
. . . 6408.96?
. . . 1.6(6)
...
...
...
...
V
( km s1) Mult
0.0
5.6
28.2
0.5
8.7
11.6
94.1
0.5
3.7
3.2
26.7
3.6
14.7
41.4
2.8
0.8
5.9
19.5
49.2
46.6
1.7
3.4
5.5
...
2.8
120.8
21.5
12.2
IC 418
IDI
ko , I/IH
...
. . .bl 5846.70
0/0 1Abl 1.3(4)
3/0
...
...
0/0
4B
...
...
...
...
1/1
7
5708.91?
...
...
1.9(5)
7/0
5C
...
0/0
3A
...
7/1
6D
...
1/1
6
7534.90
5/0
6
1.6(5)
0/0
3A
...
...
...
...
8/0
6
...
0/0
5D
...
0/0
6:
...
6/0
8
[2.8(5)]
46.6 0/0
...
...
...
...
0/0
5C
...
0/0
4A
...
0/0
4A
...
V
( km s1) Mult IDI
3.6
2.1
31.8
4.1
5.1
15.2
97.7
3.2
0.0
6.8
19.9
3.2
8.0
34.6
4.0
6.0
...
...
...
...
...
...
...
0/0
0/0
3/0
0/0
0/0
1/1
...
7/0
0/0
7/0
1/1
5/0
0/0
...
8/0
0/0
...
...
...
...
...
...
...
Ref.
5
3A
...
4C
3A
7
...
5B
4A
8
6
4A
5
...
4A
4A
...
...
...
...
...
...
...
Notes
e
Z05
e
Many
f
g
H01
E04, P04
B95
...
ZL02
h
ZL02
...
0/0
3/0
0/0
...
1/0
...
7/0
0/0
7/0
1/0
1/0
0/0
...
8/0
0/0
0/0
6/0
0/0
...
0/0
0/0
0/0
...
...
...
1/0
...
IDI
...
4A
7C
4A
...
7
...
4B
3A
5D
8
3B
5
...
4C
2A
4A
8
...
...
4A
5C
5C
...
. . .:
...
6A
...
1274
SHARPEE ET AL.
Vol. 659
TABLE 8—Continued
NGC 7027
Transition
3 SKY: 3Y2 O2 b 1þ
g YX g P12(11) k7078.58......
[ Xe vi] 5p 2P1o/2 Y5p 2P3o/2 k6408.9 ............................
He ii 5Y15 k6406.38 ..............................................
[ Fe iii] 3d 6 1S40 Y3d 6 3P22 k6408.50 ....................
C iv 9Y17 k6408.70 ...............................................
IC 418
V
ko , I/IH ( km s1) Mult IDI
...
6408.61
2.0(4)
...
...
...
13.6
104.4
5.1
4.2
...
...
...
1/0
0/0
ko , I/IH
...
...
. . .
...
. . . [9.3(6)]
4A
...
5B
...
V
( km s1) Mult IDI
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
Ref.
Notes
i
Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ‘‘OUT’’ means not in observed range; (3) ‘‘?’’ denotes an uncertain feature. Intensity: (1) numbers in
parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features.
b
Observed (ko ) transition wavelength ( km s1).
c
EMILI multiplet check statistics: number expected /number observed.
d
EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ‘‘bl’’ ( blend ). Definition of IDI given in x 6.
e
Unidentified line in IC 418 ( Sharpee et al. 2003).
f
Identified as separate line in all spectra.
g
On edge of NGC 2440, IC 2501, IC 4191 spectra, where poorest wavelength calibration is expected.
h
Listed by Hyung et al. (2001) as ‘‘unlikely ID.’’
i
IC 4191 and NGC 7027: identified as separate line.
References.—( E04) Orion Nebula ( Esteban et al. 2004); ( H01) PN IC 5217 ( Hyung et al. 2001); ( P04) PN NGC 5315 ( Peimbert et al. 2004); ( Z05) PN NGC
7027 ( Zhang et al. 2005); ( ZL02) PN Mz 3 ( Zhang & Liu 2002).
a
ion, allowing identifications of their lines to be made with more
confidence.
The line lists for the present PN sample and IC 418 (Sharpee
et al. 2003) were searched for the strongest expected Z > 30 ion
transitions within their observed bandpasses. These primarily
were the 3P1,2 Y 1D2 nebular transitions of ions with 4p2,4 and
o
o
Y 2 D3/2;5/2
nebular transitions
5p2,4 valance electrons, the 4 S3/2
3
3
of ions with 4p and 5p valence electrons, and the 2 P1o=2;3=2 Y
2 o
P1=2;3=2 fine-structure transitions for 4p, 4p5, 5p, 5p5, and 6p
valance electron ions. Fine-structure 3PY 3P transitions, when
accessible in the visible, were also included. For the cases of
o
o
Y 2 P1/2;3/2
and transauroral
Kr iv, Xe iv, and Br iv, auroral 2 D3/2;5/2
4 o
2 o
S3/2 Y P1/2;3/2 transitions were also considered. The permitted resonance lines of Ba ii and Sr ii were also searched for. Initially, all
wavelength coincidences of 1 8 or less between an observed line
and a transition wavelength were considered possible Z > 30
lines, regardless of the identifications recommended by EMILI
for those lines.
Tables 7Y10 present excerpts from the EMILI output for observed lines believed to represent the best cases for a Z > 30
transition as the actual identification for an observed line in at least
one of the PNe in which that line appeared. For each observed line
in each PN, the identifications of highest rank, as specified by the
‘‘Identification Index’’ or IDI, the EMILI figure of merit for quality of an identification (5 or less is considered a quality identification), and their multiplet search statistics (numbers expected/
observed, hereafter ‘‘multiplet statistics’’) are included. The IDI
value is followed by a letter A ! D if among the top four highest
ranked transitions, with ‘‘A’’ being the highest. Additional identifications drawn from the literature and from the terrestrial nightglow (prefaced with ‘‘SKY’’) are also included. Identifications
that did not yield predicted line intensities within 3 orders of
magnitude of the strongest value among all identifications, or
that were outside the 5 search radius, do not have a calculated
IDI value. The IDI value/rank is sometimes followed by a symbol, an asterisk indicating the most likely single identification,
‘‘bl’’ indicating components of a likely blend, and a colon indicating indeterminate alternate identifications. The reddening-corrected
intensity of the feature attributable to the putative s-process identification, if corrected for any blending as described in succeeding sections (due to other identifications denoted with a ‘‘bl’’), is
presented with a dagger below its originally observed value. The
limiting intensity, determined from the local minimum flux of
line detection considered certain (S/N ¼ 7) or from imposition
of artificial lines of S/ N ¼ 7 at that wavelength, is presented in
brackets for the case of s-process transitions without corresponding observed features. An ‘‘OUT’’ label indicates features residing outside the observed bandpass of a spectrum. Finally, features
of dubious reality are denoted by a question mark following the
observed wavelength.
Figure 4 depicts continuum-subtracted spectra, with wavelengths shifted to the nebular rest frame, in the vicinity of those
Kr, Xe, and Br lines for which a positive identification was made
in at least one PN, and which were potentially observable in all
four PNe of the present sample and in IC 418. Comments on identifications pertaining to individual Z > 30 elemental ion lines follow.
6.1. Kr Line Identifications
PB94 noted that [Kr iii] 4p4 3P2 Y4p4 1D2 k6826.70 has long
been observed in various novae and PNe but has seldom been
identified as such. They identified [ Kr iii] as a strong line in
NGC 7027, blended with weak C i ( V21) k6828.12 and He i 3s
3
SY16p 3P o k6827.88 lines.
As seen in Table 7 and Figure 5, the higher resolution of our
present PN spectra cleanly separates both the He i and C i line
from the putative [Kr iii] line, except for NGC 7027, where He i
contributes minimally to its red shoulder. However, Figure 5
shows that the R-branch head of the telluric nightglow OH Meinel
7Y3 band is a serious contaminant in this region. In NGC 7027, a
broad O vi k1032 Raman scattering line at 6829.16 8, seen previously in NGC 7027 by Zhang et al. (2005), is also observed.
The OH contribution was represented by a model of the band
normalized in intensity to the uncontaminated R1(1.5) line at
6834.01 8. The He i 3s 3SY15p 3P o k6855.91 line profile was
shifted and scaled by a factor of 0.83 to represent He i k6827.88
(Smits 1991; case B, for the most appropriate grid point: Te ¼
10 4 K, ne ¼ 10 4 cm3). The profile of the companion O vi Raman
scattering line observed at 7088 8, scaled upward by a factor of
4 to yield the best fit with the red wing of the putative [ Kr iv]
feature, represented the k6829.16 line. These line profiles and the
telluric model were subtracted from the continuum-normalized
spectra to yield the residual features shown in Figure 5.
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1275
TABLE 9
Possible Bromine Line Identifications
Transition
NGC 2440
IC 2501
V b
ko , I/IHa ( km s1) Multc IDId
V
ko , I/IH ( km s1) Mult IDI
...
[ Br iii] 4p3 4S3o/2 Y4p3 2D5o/2 k6131.0 ..................................
C iii 7h 1,3H o Y16g 1,3G k6130.30 .................................. [1.7(5)]
...
[ Ni vi] 3d 5 4D5/2 Y3d 5 2F17/2 k6130.40 .........................
[Br iii] 4p3 4S3o/2 Y4p3 2D3o/2 k6556.4 .................................. 6555.98?
Fe ii 4d 4P3/2 Y4f 2½2o5/2 k6555.94 .................................. 4.2(5)
O ii 4f F 2½4o9/2 Y6g 2[5]11/2 k6555.84 ............................
...
...
N ii 4f F 2[5/2]3 Y6d 3D2o k6556.06................................
O ii 4f F 2½4o9/2 Y6g 2[5]9/2 k6555.99..............................
...
O ii 4f 0 H 2½5o9/2;11/2 Y6g 0 2[6] kk6556.05, 6556.08.......
...
O ii 4f F 2½4o9/2 Y6g 2[4]9/2 k6556.10..............................
...
...
[ Br iv] 4p2 3P1 Y4p2 1D2 k7368.1 ......................................
C ii 3p0 2D5/2 Y3d 0 2P3o/2 k7370.00.................................. [2.6(5)]
O ii 4p 2S1o/2 Y5s 2P3/2 k7367.68 .....................................
...
C v 7p 3Po Y8d 3D 7367.60 ............................................
...
OUT
[ Br iv] 4p2 3P2 Y4p2 1D2 k9450.5 ......................................
Fe i v3D2o Y6d 2[5/2]3 k9450.95 .....................................
...
...
...
...
19.2
1.8
6.4
3.7
0.5
3.2
5.5
...
...
...
...
...
...
...
...
...
1/0
0/0
0/0
0/0
0/0
0/0
0/0
...
...
...
...
...
...
...
...
...
7:
3
3
2A
2A
2A
2A
...
...
...
...
...
...
6130.70
1.7(5)
...
6555.92?
7.3(5)
...
...
...
...
...
...
[4.4(5)]
...
...
OUT
...
NGC 7027
Transition
[ Br iii] 4p3 4S3o/2 Y4p3 2D5o/2 k6131.0 ..................................
C iii 7h 1,3Ho Y16g 1,3G k6130.30...................................
[ Ni vi] 3d 5 4D5/2 Y3d 5 2F17/2 k6130.40 .........................
[ Br iii] 4p3 4S3o/2 Y4p3 2D3o/2 k6556.4 ..................................
Fe ii 4d 4P3/2 Y4f 2½2o5/2 k6555.94 ..................................
O ii 4f F 2½4o9/2 Y6g 2[5]11/2 k6555.84 ............................
N ii 4f F 2[5/2]3 Y6d 3D2o k6556.06................................
O ii 4f F 2½4o9/2 Y6g 2[5]9/2 k6555.99..............................
O ii 4f 0 H 2½5o9/2;11/2 Y6g 0 2[6] kk6556.05, 6556.08.......
O ii 4f F 2½4o9/2 Y6g 2[4]9/2 k6556.10..............................
[ Br iv] 4p2 3P1 Y4p2 1D2 k7368.1 ......................................
C ii 3p0 2D5/2 Y3d 0 2P3o/2 k7370.00..................................
O ii 4p 2S1o/2 Y5s 2P3/2 k7367.68 .....................................
C v 7p 3Po Y8d 3D 7367.60 ............................................
[ Br iv] 4p2 3P2 Y4p2 1D2 k9450.5 ......................................
Fe i v3D2o Y6d 2[5/2]3 k9450.95 .....................................
33.3
1.0
3.9
21.5
0.5
4.1
5.9
2.7
5.5
7.8
19.5
96.9
2.4
0.8
11.1
3.2
1/1
0/0
1/0
1/1
0/0
0/0
0/0
0/0
0/0
0/0
...
...
...
...
...
...
ko , I/IH
4A 6130.34
5B 3.8(5)
7
...
6 6555.87?
2A 8.5(5)
2A
...
2A
...
3D
...
3D
...
4
...
...
...
. . . [3.5(5)]
...
...
...
...
...
OUT
...
...
V
( km s1) Mult IDI
32.3
2.0
2.9
24.2
3.2
1.4
8.7
5.5
8.2
10.5
...
...
...
...
...
...
1/1
0/0
1/0
1/1
0/0
0/0
0/0
0/0
0/0
0/0
...
...
...
...
...
...
6
2A
5D
6
2A
2A
3C
3C
3C
4
...
...
...
...
...
...
IC 418
V
ko , I/IH ( km s1) Mult
6130.32
1.1(4)
7.7(5)y
6555.93?
4.3(4)
...
...
...
...
...
7367.62?
4.4(5)
...
...
9450.85?
7.5(5)
14.7
19.6
14.7
22.0
0.9
3.7
6.4
3.2
5.9
8.2
...
...
...
...
...
...
IC 4191
...
0/0
1/0
1/0
0/0
0/0
0/0
0/0
0/0
0/0
1/0
...
1/0
0/0
1/0
0/0
IDI
V
ko , I/IH ( km s1) Mult IDI
. . .bl 6130.47
2Abl 2.8(5)
4C
...
8
...
2A [2.1(4)]
3B
...
3B
...
3B
...
4
...
5
...
9:
...
. . . [4.9(5)]
3A
...
4B
...
8:
...
3A [5.3(5)]
25.9
8.3
3.4
...
...
...
...
...
...
...
...
...
...
...
...
...
...
0/0
1/0
...
...
...
...
...
...
...
...
...
...
...
...
...
. . .
2A
4C
...
...
...
...
...
...
...
...
...
...
...
...
...
Ref.
Notes
e,f
e
B95
Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ‘‘OUT’’ means not in observed range; (3) ‘‘?’’ denotes an uncertain feature. Intensity: (1) numbers in
parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features.
b
Observed (ko ) transition wavelength ( km s1).
c
EMILI multiplet check statistics: number expected /number observed.
d
EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ‘‘bl’’ ( blend). Definition of IDI given in x 6.
e
Computed using energy levels from Biémont & Hansen (1986a); see text.
f
Unidentified line in IC 418 (Sharpee et al. 2003).
References.—( B95) PN NGC 7027 ( Baluteau et al. 1995).
a
For all PNe except NGC 2440, where the observed line appears to be entirely comprised of the OH line, there appears to be
a substantial residual feature that is coincident but, except for IC
4191, slightly to the red of the predicted wavelength for [ Kr iii]
k6826.70. The EMILI results for the corresponding observed
lines suggest that besides the C i and He i transitions mentioned
earlier, the [ Fe iv] k6826.50 feature is probably the only other
sensible alternative identification. However, given the absence
of other multiplet members that should be observed with this line
and the relative observed strengths of the likely strongest [Fe iv]
lines ( Rodrı́guez 2003), this is not a likely identification except perhaps for IC 4191. Unfortunately, the companion [ Kr iii]
3P Y 1D line at 9902.3 8, predicted to be 18 times weaker by
1
2
Biémont & Hansen (1986b), is only potentially observable in
NGC 7027, where a line listed at 9903.55 8 is much too strong
and most likely associated with C ii 4f 2F o Y5g 2G k9903.67.
Nevertheless, we conclude that the [ Kr iii] k6826.70 identification is relatively certain in IC 418, IC 2501, and NGC 7027,
uncertain in IC 4191, and that line is not observed in NGC 2440.
Examination of Table 7 shows that both transitions of the
[Kr iv] 4S o Y 2D o nebular multiplet kk5346.02, 5867.74 are clearly
identified in every spectrum, appearing as the primary EMILI
ranked identifications. For k5346.02, the [ Fe ii] k5347.65 identification is clearly too far away and the S ii (V38) k5345.71 dielectronic transition is unlikely given the absence of other multiplet
members. The Al ii k5867.8 identification for k5867.74, given by
Baldwin et al. (2000) and Sharpee et al. (2003), is clearly superseded. With a theoretical intensity ratio I(k5346.02)/I(k5867.74)
TABLE 10
Possible Identifications From Other Z > 30 Ions
NGC 2440
IC 2501
b
Transition
ko , I/IHa
V
( km s1)
Multc
IDId
ko , I/IH
[ Rb iv] 4p4 3P2 Y4p4 1D2 k5759.55......................
He ii 5Y47 k5759.74 ........................................
[ Fe ii] a2H9/2 Yc2D5/2 k5759.30 .......................
[ Rb v] 4p3 4S3o/2 Y4p3 2D3o/2 k5363.6 ....................
[ Ni iv] 3d 7 4F7/2 Y3d 7 2G9/2 k5363.35.............
O ii 4f.F 2½4o7/2 Y4d 0 2F7/2 k5363.80.................
[Cr ii] a4D3/2 Yc4D7/2 k5363.77........................
[ Te iii] 5p2 3P1 Y5p2 1D2 k7933.3.........................
He i 3p 3P o Y28s 3S kk7932.36, 7932.41.........
[ I iii] 5p2 4S3o/2 Y5p2 2D3o/2 k8536.5.......................
Cr ii 5p 6D3o/2 Y6s 6D5/2 k8536.68.....................
Ba ii 5d 2D5/2 Y6p 2P3o/2 k6141.71 ........................
O i 18d 3D2o Y4p0 3D3 k6141.75 .......................
Ne iii 4p 5P3 Y4d 5D4o k6141.48 .......................
[ Ni iii] 4s 3F2 Y4s 3P1 k6141.83.......................
[ Ba iv] 5p5 2P3o/2 Y5p5 2P1o/2 k5696.6....................
C iii 2s 1P1o Y3d 1D2 ( V2) k5695.92 ................
S ii 4d 0 2G7/2 Y5f 0 2[5]o k5696.14....................
[ Pb ii] 6p 2P1o/2 Y6p 2P3o/2 k7099.8........................
Si i 4p 1S0 Y18d (3/2, 5/2)o1 k7099.70 ..............
Ar i 3d 2½1/2o0 Y9f 2[3/2] k7099.82 .................
Ca i 4s16d 3DY3d51p 3D o k7099.82 ...............
Ca i 4s17d 1D2 Y3d47p 1P1o k7099.78..............
5759.98
5.1(4)
...
...
[3.4(5)]
...
...
OUT
...
OUT
...
6141.61
8.4(5)
...
...
...
[2.1(5)]
...
...
[2.4(5)]
...
...
...
22.4
12.5
35.4
...
...
...
...
...
...
...
...
4.9
6.8
6.4
10.7
...
...
...
...
...
...
...
...
...
0/0
0/0
...
...
...
...
...
...
...
...
...
4/0
...
5/1
...
...
...
...
...
...
...
...
...
6B
5A
...
...
...
...
...
...
...
...
. . .:
5A
...
5A:
...
...
...
...
...
...
...
...
...
[2.5(6)]
...
5363.81
6.7(6)
...
...
OUT
...
OUT
...
...
[4.0(6)]
...
...
...
[2.5(6)]
...
...
[4.5(6)]
...
...
...
NGC 7027
IC 4191
V
( km s1) Mult
...
...
...
11.7
25.7
0.6
2.2
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
4/0
0/0
6/1
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
IDI
ko , I/IH
V
( km s1)
Mult
IDI
...
...
...
. . .:
7
3A
4B
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
5759.61
2.4(5)
1.5(5)y
5363.96?
5.4(6)
...
...
OUT
...
OUT
...
6141.59
2.7(5)
...
...
...
[6.2(6)]
...
...
[6.9(6)]
...
...
...
3.1
6.8
16.1
20.1
34.1
8.9
10.6
...
...
...
...
5.9
7.8
5.4
11.7
...
...
...
...
...
...
...
...
...
0/0
0/0
...
4/0
0/0
6/0
...
...
...
...
...
4/0
...
5/0
...
...
...
...
...
...
...
...
. . .bl
4Bbl
3A
. . .:
8C
5A
7B
...
...
...
...
. . .
6A
...
7B
...
...
...
...
...
...
...
...
IDI
Ref.
Notes
IC 418
Transition
ko , I/IH
V
( km s1)
Mult
IDI
ko , I/IH
[ Rb iv] 4p4 3P2 Y4p4 1D2 k5759.55......................
He ii 5Y47 k5759.74 ........................................
[ Fe ii] a2H9/2 Yc2D5/2 k5759.30 .......................
[ Rb v] 4p3 4S3o/2 Y4p3 2D3o/2 k5363.6 ....................
[ Ni iv] 3d 7 4F7/2 Y3d7 2G9/2 k5363.35 .............
O ii 4f.F 2½4o7/2 Y4d 0 2F7/2 k5363.80.................
[Cr ii] a4D3/2 Yc4D7/2 k5363.77........................
[ Te iii] 5p2 3P1 Y5p2 1D2 k7933.3.........................
He i 3p 3Po Y28s 3S kk7932.36, 7932.41 .........
[ I iii] 5p2 4S3o/2 Y5p2 2D3o/2 k8536.5.......................
Cr ii 5p 6D3o/2 Y6s 6D5/2 k8536.68.....................
Ba ii 5d 2D5/2 Y6p 2P3o/2 k6141.71 ........................
O i 18d 3D2o Y4p0 3D3 k6141.75 .......................
Ne iii 4p 5P3 Y4d 5D4o k6141.48 .......................
[ Ni iii] 4s 3F2 Y4s 3P1 k6141.83.......................
[ Ba iv] 5p5 2P3o/2 Y5p5 2P1o/2 k5696.6....................
C iii 2s 1P1o Y3d 1D2 ( V2) k5695.92 ................
S ii 4d 0 2G7/2 Y5f 0 2[5]o k5696.14....................
[ Pb ii] 6p 2P1o/2 Y6p 2P3o/2 k7099.8........................
Si i 4p 1S0 Y18d (3/2, 5/2)o1 k7099.70 ..............
Ar i 3d 2½1/2o0 Y9f 2[3/2] k7099.82 .................
Ca i 4s16d 3DY3d51p 3D o k7099.82...............
Ca i 4s17d 1D2 Y3d47p 1P1o k7099.78..............
5759.59
3.1(4)
2.0(4)y
5363.62
9.0(5)
...
...
...
[2.0(5)]
8536.66?
1.7(5)
6141.51
7.3(5)
...
...
5696.19
5.9(5)
2.1(5)y
7099.78
4.6(5)
...
...
...
2.1
7.8
15.1
1.1
15.1
10.1
8.4
...
...
5.6
0.7
9.8
11.7
1.5
15.6
21.6
14.2
2.6
0.8
3.4
1.7
1.7
0.0
...
0/0
0/0
...
4/0
0/0
6/0
...
...
1/0
9/1
...
4/0
2/1
5/0
0/0
0/0
0/0
...
0/0
0/0
0/0
0/0
. . .bl
5Abl
6B
. . .
7B
6A
8C
...
...
6B:
4A
. . .:
8
2A:
8
8bl
6Cbl
4A
. . .
3A
4B
4B
4B
5759.78?
2.0(5)
...
...
[2.3(5)]
...
...
7932.98
1.7(5)
...
[2.3(5)]
...
[1.8(5)]
...
...
...
[1.5(5)]
...
7099.93
2.4(5)
...
...
...
V
( km s1) Mult
12.0
2.1
25.0
...
...
...
...
12.1
21.6
...
...
...
...
...
...
...
...
...
5.5
9.7
4.6
4.6
6.3
...
0/0
0/0
...
...
...
...
...
0/0
...
...
...
...
...
...
...
...
...
0/0
0/0
0/0
0/0
0/0
. . .:
3A
5B
...
...
...
...
. . .
6A
...
...
...
...
...
...
...
...
...
4A
4A
4A
4A
4A
e,f
Z05
Many
e
Z05
e
a
Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ‘‘OUT’’ means not in observed range; (3) ‘‘?’’ denotes an uncertain feature. Intensity: (1) numbers in
parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features.
b
Observed (ko ) transition wavelength ( km s1).
c
EMILI multiplet check statistics: number expected /number observed.
d
EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ‘‘bl’’ ( blend). Definition of IDI given in x 6.
e
Unidentified line in IC 418 (Sharpee et al. 2003).
f
NGC 2440, IC 4191: contaminated by flare or ghost.
References.—( Z05) PN NGC 7027 ( Zhang et al. 2005).
NEBULAR s-PROCESS ABUNDANCES
1277
Fig. 4.— Continuum-subtracted spectra in the immediate vicinity of several prominent Kr, Xe, and Br lines, shifted to each nebula’s rest frame. Predicted wavelengths
are indicated by dashed lines. Labels ‘‘NS’’ indicate telluric nightglow lines mentioned in the text. The regions around [ Kr iii] k6826.70, [ Kr v] k6256.06, and [ Xe iii]
k5846.77 are expanded in subsequent figures and discussed in the text.
of 0.65 (Schöning 1997), assuming electron densities well below
the critical density values of 1:8 ; 107 cm3 and 1:3 ; 106 cm3
for the 2D3o=2 and 2D5o=2 levels, respectively, as justified from the
values in Table 3, the observed k5346.02 intensity appears to be
somewhat too high for IC 4191 and IC 418. This suggests the
possible blending of [ Kr iii] k5346.02 with another transition,
perhaps C iii (V13.01) k5345.85, although poor multiplet statistics
and the weakness of the strongest C iii lines in IC 418 cast doubt on
this possibility for that PN. However, the other three PNe have ratios of 0.76 ( NGC 2440), 0.72 (IC 2501), and 0.73 (NGC 7027),
which are slightly higher but consistent with the expected value
within the combined measurement errors of the line intensities.
The auroral [Kr iv] 2 D3o=2; 5=2 Y 2 P3o=2 kk6107.8, 6798.4 identifications are less straightforward, except for NGC 7027 where
both are considered certain according to both PB94 and the present EMILI results. In NGC 2440, the observed line corresponding to the k6798.4 transition, which should be 2.5 times weaker
than k6107.8 ( Biémont & Hansen 1986b), is actually stronger,
while in IC 2501 the stronger k6107.8 line is not present at all.
As suggested by both PB94 and the EMILI results, C ii ( V14)
k6798.10 may be responsible for either the total intensity in IC
2501 or excess intensity in NGC 2440. While PB94 note that C ii
(V14) is of dielectronic recombination origin and its multiplet
components should not have relative intensities expected due to
LS coupling rules, an examination of the intensities of all lines
from this multiplet appearing in each PN suggests that the intensities do appear to roughly follow those rules with the exception
of what would be the strongest line at 6783.91 8. Therefore, as
an approximation, an LS-coupled relative strength of k6798.10
to k6791.47 of 0.16, instead of the one-to-one ratio with k6812.28
used by PB94, was utilized to correct the putative [Kr iv] k6798.4
feature in both NGC 2440 and NGC 7027, as k6812.28 itself appears in only one of the PN spectra (it would be the weakest
feature in the multiplet if LS coupling rules held). Subsequent
abundance analysis also suggests that the [ Kr iv] k6107.8 line
itself may be too strong relative to its nebular counterparts in
NGC 2440 and IC 4191, admitting [ Fe ii] k6107.28 as an alternate identification that appears more likely in the latter PN. In
summary, both auroral transitions can only be identified comfortably in NGC 7027.
The third magnetic dipole transition of the [Kr iv] auroral
multiplet 2 Do3=2 Y 2 P1o=2 k7131.3 is not definitively detected in any
spectrum including NGC 7027. However, in the NGC 7027
spectrum, a feature is observed at 7131.65 8 with intensity 7:7 ;
105 I( H ), comparable to k6107.8. It originally was assumed
to be an undercorrected Rowland ghost arising from the nearby
1278
SHARPEE ET AL.
Vol. 659
Fig. 4—Continued
saturated [Ar iii] 3P2 Y 1D2 k7135.80 line, but it might instead
be associated with k7131.3. The electric quadrupole line
[ Kr iv] 2 D5o=2 Y 2 P1o=2 k8091.0 is 40 times weaker and therefore
undetectable.
PB94 claim a detection of [ Kr v] 3P1 Y 1D2 k6256.06 as a blend
with C ii (V10.03) k6257.18 at their instrumental resolution. In
our spectra, as is shown in Figure 5, C ii k6257.18 is resolvable
from the putative [Kr v] feature in all spectra except NGC 7027,
where it slightly contributes to its red wing. However, in IC 2501
and NGC 7027, as also noted by Zhang et al. (2005), another C ii
line, dielectronic C ii ( V38.03) k6256.52, is believed to be a significant contaminant in some PNe based on the strength of the
nearby k6250.76 line from the same multiplet and on favorable
EMILI assessments. The profiles of both C ii lines were approximated and subtracted from all the PN spectra in Figure 5, where
the C ii (V10.03) k6259.56 line profile scaled downward by a
factor of 0.56 represents k6257.18, and the C ii k6250.76 profile,
also scaled downward by a factor of 0.56 according to LS coupling statistics, represents k6256.52. The contribution of the interloping telluric OH 9Y3 band was accounted for through subtraction
of a model of the band normalized in intensity to the nearby
Q1(2.5) k6265.21 line.
The resulting residual plots show a distinct line just to the red
of the predicted wavelength of [Kr v] k6256.06 in NGC 2440,
IC 2501, and NGC 7027, which we believe can be definitely
identified as such in all three PNe. For IC 4191, the residual flux,
3:0 ; 106 I(H ), is probably too low to be an actual line, and the
entirety of the original profile is ascribed here to C ii (V38.03)
k6256.52, particularly since both the original line and k6250.76
share the same broad profile in the spectrum. In IC 418, the putative [Kr v] profile is completely removed by the combination of
the nightglow and C ii (V38.03) k6256.52 model profiles, as is
appropriate given the nebula’s low excitation.
The other nebular line, [Kr v] 3P2 Y 1D2 k8243.39, accessible
only in the bandpasses of the NGC 7027 and IC 418 spectra, sits
amid the head of the Paschen series, rendering it difficult to disentangle at the resolution of our NGC 7027 spectrum. In the IC 418
spectrum, a comparable observed line is indistinguishable from
other nearby lines in the Paschen series and is clearly identifiable
as H i 3Y43 k8243.39, with the Kr v identification unlikely due
to the PN’s low ionization level. Returning to NGC 7027, PB94
identified this transition at 8242.7 8 as a blend with N i (V2)
k8242.39 and O iii 5g G 2[9/2]o Y6h H 2[11/2] k8244.10, compromised by telluric absorption. In our NGC 7027 spectrum, [Kr v]
k8243.39 is tentatively identified as the blue peak at 8244.33 8 of
a resolved two-line blend with H i 3Y42 k8245.64, which appears significantly affected by telluric absorption. The alternate identification for the line, N i k8242.39, is clearly resolved
here as a separate feature. Because this observed line has a
measured intensity ratio with k6256.06 roughly comparable to
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1279
Fig. 4—Continued
what would be expected if both lines were due to [ Kr v],
I(k6256:06)/I(k8243:39) ¼ 1:1 ( Biémont & Hansen 1986a), it
is believed that O iii k8244.10 contributes only a minor amount
to the line, although the presence of the telluric absorption complicates matters. To summarize, both [ Kr v] lines appear to be
present with about the expected intensity ratio in the spectrum
of NGC 7027, but both are absent in the spectrum of IC 418. For
the cases of IC 2501 and NGC 2440, [ Kr v] k6256.06 is probably present.
The spectra were also examined for evidence of other auroral and transauroral transitions of Kr ions. Only the auroral line
[ Kr v] 1D2 Y 1S0 k5131.78 satisfied the initial 1 8 screening
criterion and also appeared in the EMILI output as a possible
identification for an observed line at 5131.0 8 in all PNe.
The case for this identification is lessened by the expected weakness of this line compared to other [Kr v] lines that were not definitively identified, and particularly because of its high observed
intensity of 1:3 ; 103 I(H) relative to the 1 ; 104 I(H ) for
kk6256.06, 8243.39 in NGC 7027, the nebula that might be expected to exhibit the best evidence for this line given the strength
of the other confirmed Kr lines. Instead, either O i 3p 3PY8d 3D o
k5131.25, which was the highest ranked line in all but one nebula
(IC 418) and has a strength comparable to other members of the
same sequence, or C iii 5g 3GY7h 3H o k5130.83 appears the more
likely identification.
6.2. Xe Line Identifications
A number of Xe ion transition identifications are also considered probable in our spectra. The EMILI statistics for Xe identifications are affected by the low solar Xe abundance, which is an
order of magnitude lower than Kr (Lodders 2003). This contributes to low predicted relative emission-line intensities and consequently higher or absent IDI values for its identifications as
compared to those computed for weak transitions from more abundant elements. As such, every appearance of a Xe ion transition
as a candidate identification for an observed line in an EMILI list,
regardless of its IDI value, was given serious consideration as it
signaled that alternative identifications from more abundant elements did not predominate despite the advantage arising from
greater abundances. Table 8 lists the EMILI statistics for lines
judged most likely to correspond to Xe ion transitions, while
Figure 4 depicts the regions around the most likely observed
transitions.
The identification of [ Xe iii] 3P2 Y 1D2 k5846.77 in NGC 7027
was given by PB94 to an excess intensity in the He ii 5Y31
k5846.66 line. This excess was searched for in the present PN
sample through the subtraction of the He ii k5837.06 profiles,
shifted to the k5846.66 line position and scaled assuming
I(k5837:06)/I(k5846:66) ¼ 0:92 (Storey & Hummer 1995;
case B, for the most appropriate grid point: Te ¼ 10 4 K,
1280
SHARPEE ET AL.
Vol. 659
Fig. 4—Continued
ne ¼ 10 4 cm3), as is shown in Figure 6. The intensity of He ii
k5837.06 was not corrected for the presence of C iii 7h 3H o Y
18i 3I k5836.70 because the line used for this correction by PB94,
C iii 7i 3IY18k 3K o k5841.2, was either not present or an identification of low rank ( IDI ¼ 6, ranked fifth) for a corresponding
line in NGC 7027. As seen in Figure 6, distinct residual profiles are present after subtraction of He ii k5846.66 for both
NGC 7027 and IC 418 (the latter having negligible He ii) and at
the correct wavelength for [ Xe iii] k5846.77. A bizarrely shaped
profile is seen for IC 2501, and no profiles are seen for either
NGC 2440 or IC 4191 after subtraction. While the corresponding
observed lines are probably He ii k5846.66 in the latter two PNe,
IC 2501 has very weak He ii lines, suggesting that the residual profile is attributable to something else, tentatively [Xe iii]
k5846.77. The [ Xe iii] k5846.77 identification, while not appearing in the EMILI lists for NGC 7027 or IC 2501, is considered a
better choice than the other top ranked identification, [Fe ii]
k5847.32, which has poorer wavelength agreement and multiplet statistics in all cases. While the [ Xe iii] 3P1 Y 1D2 transition
at 1.37 m is unavailable for confirmation, we believe that the
[ Xe iii] k5846.77 is definitely present in NGC 7027 and IC 418
but is only tentatively identified in IC 2501.
The reality of the observed features potentially corresponding to the [ Xe iv] 4 S 5o=2; 3=2 Y 2 D3o=2 kk5709.2, 7535.4 transitions is
marginal except in NGC 7027. However, their identifications as
the Xe lines, should they be actual emission lines, are more certain. For k5709.2, the widely observed N ii (V3) k5710.77 line is
detected as a separate line in all of the spectra. Alternative identifications such as [Fe i] a5D3 Ya5P1 k5708.97 either have too
many missing multiplet members, except for IC 2501 where the
ratio of potentially observed to total number of multiplet lines expected is marginally better, or, as in the case of the Fe ii] k5709.04
intercombination line, are unlikely given that permitted Fe ii lines
were not anywhere definitively identified. For k7535.2, the reality
of the observed lines is less uncertain, including IC 418 where a
previously unreported line has been uncovered by more thorough
examination of its spectra. The Fe ii] (V87) k7534.82 identification proposed by Hyung et al. (2001) is unlikely for the
same reasons as Fe ii] k5709.04 for [ Xe iv] k5709.2. The N ii
5f G 2[7/2]4 Y10d 1F3o k7535.10 identification, selected by Esteban
et al. (2004) and Peimbert et al. (2004), also appears unlikely
given the relative strengths of other known permitted N ii lines
in the spectra. It should be noted that the EMILI statistics for
the k7535.10 identifications in the MIKE PN sample would
probably have been better if the line did not appear near the
end of the last spectral order where the wavelength calibration is the poorest. Assuming an electron density well below the
critical regime (107 cm3), the expected intensity ratio is
I(k5709:2)/I(k7535:4) ¼ 0:64 (Schöning & Butler 1998). The
observed range of values that we observe, 0.33Y1.19, suggests
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1281
Fig. 4—Continued
measurement errors on par with the average values for such weak
lines. The strongest examples in NGC 7027 do yield the bestagreeing ratio (0.82). As such, we conclude that both [Xe iv] lines
may be present in all PN spectra with varying degrees of certainty.
One of the stated goals for future spectroscopy of NGC 7027
and other PNe given by PB94 was the detection of [Xe v] transitions, none of which they detected with any certainty. In the
present spectra, no observed line passed the 1 8 initial selection criterion for the 3P1, 2 Y 1D2 kk5228.8, 6998.7 lines, nor did
these transitions appear as possible EMILI IDs for any observed
line. This is despite the improved sensitivity and resolution that
should have enhanced the chances of detecting k5228.8 and that
should have easily separated k6998.7 from O i ( V21) k7002.10,
the contaminant noted by PB94. Nevertheless, the detection of
these lines in our spectra would remain problematic since for
k5228.8 extensive flaring from [O iii] k5006.84 in an adjacent
order leads to numerous ghosts in its vicinity, while for k6998.7
telluric absorption in the tail end of the Fraunhofer A band
complicates its observability.
EMILI did suggest the fine-structure transition [ Xe v] 3P0 Y 3P2
k7076.8 as a possible identification for observed lines in two
PNe, IC 4191 and NGC 7027, PNe with appropriate excitation
levels for the appearance of a line of this ion. This is the primary
EMILI identification in IC 4191. A third occurrence of the corresponding observed line in NGC 2440 is clearly associated with
3 a nightglow line from the O2 b 1 þ
g YX g 3Y2 band on comparison with a simulation of that band. The competing identification C i (V26.01) k7076.48 has poor multiplet statistics, while
[ Fe iii] k7078.10 and [Ni ii] k7078.04 are feasible but have poor
wavelength agreement. The Ca i identifications are also unlikely
given that they arise from energy levels of large energy, where
lines that would follow from cascades from these levels are not
observed. Two remaining obstacles to an identification with
[ Xe v] k7076.8 are the lack of the nebular 3PY 1D lines and the
low branching ratio for this electric quadrupole transition (0.008
with respect to 3P1 Y 3P2 at 2.07 m). Nevertheless, despite this
transition’s expected weakness, the lack of viable alternate identifications suggests that the corresponding observed line may be
at least tentatively identified in NGC 7027 and probably identified in IC 4191, and both occurrences yield reasonable abundance values in subsequent analysis.
The fine-structure transition [ Xe vi] 2 P1o=2 Y 2 P3o=2 k6408.9 was
identified with certainty in NGC 7027 by PB94. In the present
NGC 7027 spectrum, k6408.9 is probably associated with a weak
but clearly separable observed line on the red wing of He ii 5Y15
k6406.38. The alternate primary EMILI identification of [Fe iii]
k6408.50 was not considered likely, as an inspection of the NGC
7027 spectrum for other lines from all low-energy [Fe iii] multiplets did not show a significant number of matches to warrant
strong consideration as a likely identification. [ Fe iii] k6408.50
1282
SHARPEE ET AL.
Vol. 659
Fig. 4—Continued
also has a comparatively high excitation energy (6.25 eV ). A
C iv 9Y17 k6408.70 identification is similarly downgraded by
its high excitation. The reality of corresponding observed lines
in NGC 2440 and IC 4191 is questionable, although for lack of
suitable alternate identifications, uncertain identifications of
[ Xe vi] k6408.90 are retained for both.
As with Kr, an inspection for auroral and transauroral transitions of various Xe ions was also undertaken. There were coincidences within 1 8 between observed lines and the [Xe iii]
3
P1 Y 1S0 k3799.96, [Xe iii] 1D2 Y 1S0 k5260.53, [Xe iv] 4 S3o=2 Y
2 o
P1=2 k3565.8, [ Xe iv] 2 D3o=2 Y 2 P3o=2 k4466.5, [ Xe iv] 2 D5o=2 Y
2 P o k5511.5, [Xe iv] 2 D o Y 2 P o k6768.9, and [Xe v] 1D2 Y 1S
0
3=2
3=2
1=2
k6225.3 transitions. However, except for two instances, the
identifications did not appear in the EMILI list for those particular lines, and most had more reasonable and higher ranked
identifications: C ii ( V30) kk5259.66, 5259.76, for example,
instead of the auroral [ Xe iv] k5260.53 line. In many cases the
line only appeared in one of the PN spectra analyzed here and
could be rejected due to the strength or absence of the nebular
transitions from the same ions.
6.3. Br Line Identifications
In PB94 two transitions of Br were identified: [Br iii] 4 S3o=2 Y
D5o=2 k6131.0 with certainty, and [Br iv] 2 D3o=2 Y 2 P3o=2 k7385.1
as possible. A second nebular transition, [ Br iii] 4 S3o=2 Y 2 D5o=2
2
k6556.4, was lost in a blend amid the [N ii] k6548.04+H+[ N ii]
k6583.46 complex.
The source for the Br iii levels in the Atomic Line List version
2.05 is Moore (1958), while the source used by PB94 is stated
to be the experimental levels listed in Table IV of Biémont &
Hansen (1986a), from unpublished work of Y. N. Joshi and
Th. A. M. van Kleef, and provided by private communication with
van Kleef. However, a comparison of the Ritz-determined wavelengths from both sets of levels with those listed by PB94 shows
that the Joshi and van Kleef 2Do term energy levels were used for
the nebular transition wavelengths, while the Moore (1958) levels were used for auroral transition wavelengths. The substantial
difference in the 2D3o=2 level energy, 15,042 cm1 for Moore
(1958) and 15,248 cm1 for Joshi and van Kleef, leads to a difference in the 4 S3o=2 Y 2 D3o=2 transition air wavelength of 6646.3 8
versus 6556.4 8, respectively, with a lesser difference for the other
transition, 6132.9 8 versus 6131.0 8. Since the Moore (1958)
levels used by the Atomic Line List version 2.05 and therefore by
EMILI have stated uncertainties of 0.63 cm1, while those for
Joshi and van Kleef in Table IV of Biémont & Hansen (1986a)
are listed to a precision of 1 cm1, nearly the same level of uncertainty, it is difficult to know which levels are more accurate.
Therefore, EMILI was run using both sets of energy levels, and
the 4S o Y 2D o lines were searched for at both sets of resultant
wavelengths.
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1283
Fig. 4—Continued
For the transition wavelengths generated from the Moore
(1958) energy levels, kk6132.9 and 6646.3, no observed line was
detected in any spectrum meeting the initial 1 8 selection criterion. However, for the 4 S3o=2 Y 2 D5o=2 transition wavelength generated from those energy levels used by PB94, k6131.0, there do
appear to be observed lines in four of the five PN spectra near the
k6130.4 line that PB94 identified with [ Br iii].
Although [Br iii] k6131.0 is the highest ranked EMILI identification in only one PN, as seen in Table 9, other identifications
are not compelling. PB94 cite their observed line as being a
blend with C iii 7h 1,3H o Y16g 1,3G k6130.30, which is the highest
EMILI-ranked transition in each spectrum in which the putative
Br iii line is observed, except for IC 2501. However, the companion line C iii 7h 1,3H o Y16i 1,3I k6126.30, claimed by PB94 to
have an equal intensity to k6130.30, is not present in either the
IC 4191 or IC 418 spectra. Well-known lower excitation C iii
lines are of negligible intensity in IC 418, while in IC 2501 C iii
k6126.30 has poor wavelength agreement with its corresponding
observed line and is not the primary EMILI identification for that
line. Only in NGC 7027 is C iii k6126.30 a primary ID, suggesting that only in its spectrum is the equally intense C iii k6130.30
likely to be present and accountable for at least a portion of the
putative [Br iii] k6131.0 line. As such for NGC 7027, the k6126.30
line intensity is subtracted from the observed line at 6130.32 8,
while in the remaining spectra the corresponding observed lines
are ascribed wholly to [Br iii] k6131.0. Possible contamination
due to flaring from an adjacent order in the NGC 7027 spectrum
was not accounted for. The [ Ni vi] k6130.40 identification, the
second ranked identification in many of the spectra, is of high
excitation (8.3 eV upper level), too high an ionization for IC 418,
and appears only to be ranked highly due to a favorable coincidence in wavelength.
Spectroscopic confirmation for this identification is sought in
the possible presence of the [Br iii] 4 S3o=2 Y 2 D3o=2 k6556.4 transition, in the PN spectra at expected intensities similar to that of
k6131.0. As mentioned, detection of this line is difficult given its
proximity to the saturated H line and its associated ghosts in
its immediate proximity. Nevertheless, inspection of the line lists
and spectra (Fig. 4) does show distinct co-aligned features in the
IC 2501, IC 4191, and NGC 7027 spectra near this wavelength.
While initially assumed to be ghosts, the fact that these features
can be found in spectra from two different instruments and appear invariant to the H intensity suggests that they may be unrelated to H. That real lines can be observed between [ N ii]
k6548 and H is demonstrated by the presence of the strong
telluric OH P1(3.5) 6Y1 k6553.62 line just to the blue of the suspected [Br iii] feature. EMILI lists the [Br iii] k6556.4 identification among possible choices in all PNe in which the line appears,
with the k6131.0 identification satisfying the multiplet search in
IC 2501 and IC 4191. Alternate EMILI-favored identifications
1284
SHARPEE ET AL.
Vol. 659
Fig. 5.—Left: Spectra in the vicinity of [ Kr iii] k6826.70 (black solid line) shifted to nebular rest wavelengths, with similarly shifted and superimposed co-added
modeled spectra of telluric emission and observed interloping nebular features ( gray dot-dashed line). Plus signs indicate telluric OH Meinel 7Y2 R1(3.5) k6827.46,
R1(4.5) k6828.47, and R1(2.5) k6829.49 lines (left to right), nebular He i k6827.88 is marked with an arrow, and an O vi k1032 Raman feature at 6829.16 8 ( NGC 7027)
is marked with a lower arrow. Right: Same as the left panel, but for [ Kr v] k6256.06, with OH Meinel 9Y3 Q2(0.5) k6256.94 and Q1(1.5) k6257.96 marked as plus signs (left
to right) and nebular C ii kk6256.52, 6257.18, and 6259.65 lines marked with arrows (left to right). Panels to the right show residuals after subtraction in the vicinity of
predicted wavelengths of the target line (vertical dashed line in all panels).
for these features correspond to O ii and N ii permitted and coreexcited transitions between levels of primary quantum number
4Y6 and are doubtful as they are of comparatively high excitation
with respect to better known 3Y3 or 3Y4 transitions in these PNe
that appear as lines of equal or lesser intensities. They appear here
only due to their wavelength agreement with the observed lines.
The highest ranked Fe ii k6555.94 is discounted once again by the
lack of other permitted Fe transitions in any spectra.
The observed intensity ratios of the two putative [Br iii] lines,
I(k6131.0)/I(k6556.4), ranging from 0.18 to 0.45, do conflict
with theoretical expectation as follows. While specific collision
strengths are unavailable at present for the Br iii levels, the ratio
of the strengths relevant to these transitions should still be roughly
proportional to their respective upper level statistical weights
( Péquignot & Baluteau 1994; Osterbrock & Ferland 2006),
weighted by a Boltzmann factor respecting the difference in level
energies, and nonnegligible fine-structure emission between the
2 o
D levels, at subcritical electron densities. Employing the IRAF
NEBULAR package task ionic to solve for the relative populations of the 2D o levels at temperature derived from the diagnostic
thought to be the most appropriate, from [Ar iii], and using appropriate Kr iv collision strengths as a proxy, scaled as discussed
in x 7, the I(k6131.0)/I(k6556.4) ratio is expected to be 0.72Y
0.83. This is close to the 0.67 value expected from the ratio of the
collision strengths alone and 2Y3 times larger than the ratio observed in the spectra. At these levels the [Br iii] k6131.0 line should
have been observable in NGC 2440, as the predicted intensity ex-
ceeds the likely detection limit for features in its vicinity. Therefore, we consider [Br iii] to have been only tentatively detected in
NGC 2440. However, at least one [ Br iii] line does appear to be
present in the remaining four PNe. They all have lines detected at
the same wavelengths, although shifted 20Y30 km s1 to the blue
of the wavelengths we have adopted for the [Br iii] lines, at approximately the same relative intensities, and they lack satisfying alternate identifications. The ratios of observed intensities (or
in the case of IC 418 the upper limit), while not exact, are within
a factor of 2Y3 of those expected.
The identifications of the nebular [Br iii] lines suggest that the
level energies of Moore (1958) for at least the 2D o term levels
may be in error. PB94 also identified an observed line at 7384.3 8
as possibly auroral 2 D3o=2 Y 2 P3o=2 k7385.2, but no such line appears in any of the PN spectra examined here. However, since
PB94 appears to have used 2P o Y 2D o transition wavelengths
derived from the Moore (1958) energy levels, this might not
be a surprise, although no lines in any PN were discovered
at the corresponding wavelength for the 2 D3o=2 Y 2 P3o=2 transition
(7483.1 8) computed from the levels listed by Biémont &
Hansen (1986a) either.
The nebular [ Br iv] 3P1 Y 1D2 k7368.0 line was detected by
PB94 at 7366.0 8, affected by telluric absorption, and blended
with C iv 10Y21 k7363.9 and O iii 4s 3P1o Y4p 3P1 k7365.35. In our
NGC 7027 spectrum, a line appearing at 7367.62 8 as a small
protrusion above the local continuum level interpolated between
telluric absorption features in both spectral orders covering that
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1285
may be spurious. No auroral or transauroral features of [Br iv]
were identified in any spectrum.
6.4. Other Z > 30 Line Identifications
Fig. 6.—Spectra in the vicinity of [ Xe iii] k5846.77 (black solid line) shifted
to nebular rest wavelengths, with the superimposed profile of the He ii 5Y32
k5837.06 shifted to the rest wavelength of the He ii 5Y31 line and scaled as described in the text ( gray dot-dashed line). Panels to the right show residual spectra
after subtraction near the rest wavelength of the [ Xe iii] k5846.77 line (vertical
dashed line in all panels).
wavelength is tentatively identified as [ Br iv] k7368.0. The
alternate candidate listed by PB94, dielectronic C ii 3p0 2D5/2 Y3d 0
2 o
P3=2 k7377.00, is too far away, while an O ii k7367.68 identification
is doubtful given its high upper level energy. The C v 7p 3P o Y
8d 3D k7367.60 transition is a possible alternate identification,
but only one other C v transition in the spectrum, C v 6gh 1,3G,H o Y
7hi 1,3H o, I k4944.50 with an IDI value of 3, is a top ranked
EMILI IDI; others are of lesser rank or do not appear in the
EMILI lists for the corresponding observed lines.
The other nebular transition, [Br iv] 3P2 Y 1D2 k9450.5, was not
detected by PB94, even though the line should be of the same
intensity ( Biémont & Hansen 1986a). Comparison between the
IC 418 and NGC 7027 spectra shows that something is filling in
the telluric absorption feature in the latter at 9450.85 8, with an
estimated intensity close to [Br iv] k7368.10, although a ghost
feature at this wavelength arising from scattered light within the
spectrograph cannot be ruled out based on the proximity of other
similar features. If the feature is real, however, the alternate identification of Fe i k9450.95 is not compelling. Therefore, both
transitions are identified in NGC 7027, although tentatively since
they appear in only one PN spectra, they may be attributable to
misinterpretation of the local continuum level complicated by
telluric absorption of varying degree, and the latter observed line
The spectra were searched for lines originating from other
Z > 30 ions. Among the numerous coincidences between observed lines and transition wavelengths at the 1 8 level, those
listed in Table 10 were judged to be the most likely to correspond
to real lines describable by a Z > 30 identification in at least
one of the PNe. Given the low solar abundances of these ions,
only transitions among the lowest lying levels were expected to
be observable.
PB94 declared as certain the identification of a line at 5758.7 8
in NGC 7027 as [Rb iv] 3P2 Y 1D2 k5759.44. Corresponding lines
are detected in four of five PNe here. However, at even the highest instrumental resolution, this transition would blend with He ii
5Y47 k5759.74. In NGC 2440 and IC 4191 there is contamination from a flare or ghost at the position of He ii 5Y47 k5759.74 in
one spectral order. Using only the uncontaminated order, and inspecting the nearby He ii 5Yn sequence, suggests that in IC 4191
the He ii k5759.74 line shows significant excess, while in NGC
2440 there is better agreement with recombination theory and the
line is likely He ii. The NGC 7027 spectrum, taken with a different
instrument, does not show any contamination but do show a similar
excess in He ii 5Y47 k5759.74. Thus, the line intensities in IC
4191 and NGC 7027 were corrected by I(5Y47)/I(5Y46) ¼ 0:95
derived from the ratio of their emissivities (Storey & Hummer
1995). IC 418 has no detectable He ii lines, enhancing the [Rb iv]
k5759.44 identification, but the weakness and irregularity of its
profile cast doubt on its reality as a line, and it is only tentatively
identified here. The alternate EMILI recommended identification, [ Fe ii] k5759.30, was not considered likely given its high
excitation energy. Since the expected intensity ratio of 3P2 Y
1
D2 k5759.44 to 3P1 Y 1D2 k9008.75 is 17 ( Biémont & Hansen
1986b), the apparition of k9008.75 in IC 418 and NGC 7027 is
probably likely due to He i 3d 3DY10p 3P o kk9009.23, 9009.26
or is unreal. In summary, the [Rb iv] k5759.55 line is identified in
IC 4191, NGC 7027, and tentatively in IC 418.
Another Rb line detection considered probable by PB94 in
NGC 7027 is [Rb v] 4 S3o=2 Y 2 D3o=2 k5363.6, which they identified
at 5364.2 8. Although it does not appear in the EMILI list for the
corresponding lines in our spectra, we believe that its identification in at least NGC 7027 is viable given the poor multiplet
statistics of [ Ni iv] k5363.35, where none of the four other multiplet lines are clearly present. The O ii 4f F 2½4o7=2 Y4d 0 2F7/2
k5363.80 identification is an interesting alternative. Well-known
O ii dielectronic doublet lines of multiplets V15, V16, and V36,
such as 3s 0 2D5/2 Y3p0 2F7o=2 (V15) k4590.97, are all clearly present in all of our PN spectra, except NGC 7027 where they are
either outside the bandpass or not optimally placed for detection,
at intensities 104I(H). These lines are present at similar intensities in the NGC 7027 spectrum of Zhang et al. (2005). Lines
from 3dY4f transitions, particularly 3d 2D5/2 Y 4f F 2½4o7=2 (V92a)
k4609.4, arising from the lower level of the O ii k5363.80 transition are also favorably identified at roughly the same intensity.
Therefore, it is not out of the realm of possibility that this line is
evidence of the partial feeding of the 2½4o7=2 level. However,
while EMILI did not perform a multiplet check on this line (due
to a nonYLS-coupled lower level), two other transitions at 5361.74
and 5375.57 8 from the same upper term and ending on this level
are not present in any of the spectra. Without spontaneous emission coefficients it is difficult to judge a potential branching ratio
for these transitions. Therefore, it is believed that [Rb iv] k5363.60
1286
SHARPEE ET AL.
might be tentatively identified in IC 2501 and IC 4191 as well.
Unfortunately, the matching 4 S3o=2 Y 2 Do5=2 k4742.40 line is disguised by a flare or ghost in the MIKE spectra and is not optimally
placed in NGC 7027 to allow a spectroscopic confirmation, nor
observed in IC 2501 and IC 4191.
Numerous lines of various Sr ions were searched for, including the strong [I /I(H) ¼ 3 ; 104 ] [Sr iv] 2 P3o=2 Y 2 P1o=2 k10276.9
fine-structure line named as a possible identification by PB94
o
Y
in NGC 7027, and the [Sr vi] 4 S3o=2 Y 2 D5o=2 k4249.2 and 2 D3/2
2 o
P1=2 k5434.3 lines identified as possible and tentative, respectively, by PB94 and also identified by Zhang et al. (2005).
The Sr ii kk4077.71, 4215.52 resonance lines and the [Sr ii]
2 S = Y 2 D = = kk6738.39, 6868.17 lines observed in Carinae
12
5 2;3 2
( Zethson et al. 2001) were also sought. While there were some
instances of wavelength coincidences with observed lines, the
magnitude of the wavelength differences and existence of plausible alternate identifications, such as [Fe ii] 4249.08 for [Sr vi]
k4249.2, did not warrant a claim of identification. The same was
o
Y 2 P1o=2 k8023.6 and [Zr vii]
true for Yand Zr, where the [Y v] 2 P3/2
3
3
P2 Y P0 k7961.4 fine-structure lines, listed as possible and tentative identifications by PB94, respectively, did not have any
wavelength coincidence with an observed line in any PNe within
the initial 1 8 screening criteria.
Turning to the fifth row of the periodic table, we believe that
an observed and previously unidentified line in IC 418 may
correspond to [ Te iii] 3P1 Y 1D2 k7933.3, which would be the first
visual identification of a Te iii line in a PN. The He i 3p 3P oYns 3S
sequence of lines, of which the alternate identifications He i
kk7932.36, 7932.41 are members, does not become clearly evident in this spectrum until 3pY10s at 8632.76 and 8632.83 8.
The wavelength coverage of our spectra does not extend out to
the possibly stronger 3P2 Y 1D2 k10876.0 transition, so this line
cannot be used to check the Te iii identification. Similarly, the deo
o
Y 2 D3/2
k8536.6 line in NGC 7027,
tection of a possible [I iii] 4 S3/2
with a less than compelling Cr ii permitted line as an alternate
primary EMILI identification, cannot be confirmed through obo
o
Y 2 D5/2
k6708.7 transition. The
servation of its companion 4 S3/2
putative k6708.7 line appears to be better explained by a combination of [Mn ii] k6709.93 and possibly [Cr v] k6709.8, although
the latter line has poor multiplet statistics (2/0). Nevertheless, the
[I iii] k8536.6 transition has a definite IDI that is second ranked
for its corresponding observed line, so we count this as a tentative
detection for NGC 7027.
PB94 claims the detection of four transitions belonging to permitted multiplets of Ba ii, with three detections considered certain and one considered possible. Some of these same transitions
have also been identified in emission by Zhang et al. (2005) in
NGC 7027 and by Hobbs et al. (2004) in the compact H ii region within the Red Rectangle. In the present sample it is believed that one of the transitions considered certain by PB94,
o
k6141.71, may correspond to an obBa ii 5d 2D5/2 Y6p 2P3/2
served line in three of five PNe. Given the low solar abundance
and high condensation temperature (1455 K; Lodders 2003) of
Ba, the detection of these transitions might initially be considered unlikely. However, as originally proposed by PB94, if
sufficient gas-phase Ba+ is available for collisional excitation,
emission, and moderate self-absorption, it is estimated that
o
k4554.03
optical depths of 0.5 and 2.7 in the 6s 2S1/2 Y6p 2P3/2
resonance transition are sufficient to account for both its apparent
nondetection and the observed intensity of the putative k6141.71
lines in NGC 2440 and IC 4191, respectively. PB94 estimated an
optical depth of 3 for the k4554.03 transition in NGC 7027. Under these scenarios the k6141.71 line is either among the strongest or is the strongest observable Ba ii line, with its relative
Vol. 659
intensity with respect to other Ba ii lines increasing with a larger
optical depth in the k4554.03 line.
The EMILI results do suggest some potentially viable alternative identifications other than Ba ii for individual PNe, but none
that can satisfactorily account for the observed line in all three.
O i 18d 3Do2 Y4p0 3D3 k6141.75 arises from an autoionizing level,
is too strong with respect to other permitted O i transitions further
down the cascade chain, and is not accompanied by any other
multiplet members in any PNe. Under the abundance and ionization model created by EMILI for NGC 2440 and IC 4191, Ne iii
4p 5P3 Y 4d 5D4o k6141.48 did not produce an emission line with
an intensity within 3 orders of magnitude of the strongest predicted intensity among all putative identifications. This was not
the case in NGC 7027, where the line was among the strongest
predicted lines and is expected to be among the strongest in the
multiplet, and where the identification is enhanced by favorable
multiplet statistics. The [ Ni iii] 4s 3F2 Y 4s 3P1 k6141.83 arises
from a level of high-excitation energy (9.9 eV ) and is probably
not the strongest member of its multiplet as it does not originate
from the level of the multiplet with the highest statistical weight,
and other potentially stronger multiplet members are not observed in IC 2501 and NGC 7027. Yet in NGC 2440, the multiplet statistics are somewhat better, and the large intensity of the
corresponding observed line and the Ba abundance derived from
it is at odds with lower abundances derived for Kr and Xe in subsequent abundance analysis. The excellent agreement between
the observed wavelengths in all three PNe and the Ba ii k6141.71
wavelength, the expectation that k6141.71 is the strongest Ba ii
line, and the lack of a good alternate identification for IC 4191
suggest that Ba ii k6141.71 warrants serious consideration as the
correct identification in all cases.
o
o
Y 2 P1/2
k5696.6,
The fine-structure transition [Ba iv] 2 P3/2
another certain detection from PB94, is verified by its detection in our NGC 7027 spectrum after a correction for C iii (V2)
k5695.92 is made. The intensity attributable to k5696.6 was determined using the effective recombination coefficient for C iii
k5695.92 ( ¼ 3:1 ; 1015 cm3 s1) specified by PB94, those
from Nussbaumer & Storey (1984) and Péquignot et al. (1991)
for C iii 5g 1,3GY6h 1,3H o k8196.61, and the k8196.61 line’s observed intensity. Subtraction yielded a line amounting to 36% of
the originally observed intensity.
The only sixth-row elemental transition sought in these speco
o
Y 2 P3/2
k7099.80 line, which
tra was the fine-structure [Pb ii] 2 P1/2
was identified with certainty by PB94. The transition appears in
the EMILI list for IC 418 tied for the highest ranked transition for
a previously unidentified line. For NGC 7027, the transition is
not ranked, but the competing identifications in its EMILI list are
not convincing given their high excitation energies. Excellent
wavelength agreement is seen in both cases, and both identifications are considered certain.
In summary, of the 18 Z > 30 elemental ion transitions considered certain or probable by PB94 in NGC 7027, 15 are believed to be detected to various degrees of certainty in the present
set of spectra. The final intensities of all Z > 30 lines detected
with any certainty within at least one PN are presented in Table 11,
with certain identification shown with an asterisk and tentative
identifications without. Conspicuous among those missing from
PB94 is [Se iii] 3P1 Y 1D2 k8854.2. The nearby He i 3d 3DY11p 3P o
k8854.14 transition does appear to show a large excess relative to that expected from other confirmed lines in the same series
in both IC 418 and IC 7027. However, the energy levels in the
Atomic Line List version 2.05, derived from Moore (1952) and
also utilized by PB94, have a listed uncertainty of 6.3 cm1 that
exceeds the maximum 1 cm1 tolerance allowed for inclusion of
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
1287
TABLE 11
Summary of Possible Z > 30 Ion Line Intensities
PNe
H ii Regions
NGC 7027
Transition
NGC 2440
IC 2501
IC 4191
(a)
[ Br iii] k6131.0 .....................
[ Br iii] k6556.4 .....................
[ Br iv] k7368.1.....................
[ Br iv] k9450.5.....................
[ Kr iii] k6826.70...................
[ Kr iv] k5346.02...................
[ Kr iv] k5867.74...................
[ Kr iv] k6107.8.....................
[ Kr iv] k6798.4.....................
[ Kr v] k6256.06....................
[ Kr v] k8243.39....................
[ Rb iv] k5759.55 ..................
[ Rb v] k5363.6 .....................
[ Te iii] k7933.3 .....................
[ I iii] k8536.5........................
[ Xe iii] k5846.77 ..................
[ Xe iv] k5709.2 ....................
[ Xe iv] k7535.4 ....................
[ Xe v] k7076.8 .....................
[ Xe vi] k6408.9 ....................
Ba ii k6141.71.......................
[ Ba iv] k5696.6 ....................
[ Pb ii] k7099.8......................
<1.7(5)
4.2(5)
<2.6(5)
...
...
3.5(4)
4.6(4)
5.6(6)
1.3(5)
1.5(5)
...
...
<3.4(5)
...
...
...
1.2(5)
1.3(5)
...
2.8(5)
8.4(5)
<2.1(5)
<2.4(5)
1.7(5)
7.3(5)
<4.4(5)
...
2.5(5)
6.1(5)
8.5(5)
<4.3(6)
...
8.0(6)
...
<2.5(6)
6.7(6)
...
...
7.7(6)
6.5(6)
1.6(5)
<5.5(5)
<2.4(6)
<4.0(6)
<2.5(6)
<4.5(6)
3.8(5)
8.5(5)
<3.5(5)
...
1.5(5)
2.4(4)
1.9(4)
...
<8.1(6)
...
...
1.5(5)
5.4(6)
...
...
...
1.2(5)
3.6(5)
3.9(6)
1.6(6)
2.7(5)
<6.2(6)
<6.9(6)
7.7(5)
4.3(4)
4.4(5)
7.5(5)
4.6(4)
1.9(3)
2.6(3)
7.2(5)
1.6(5)
1.2(4)
1.4(4)
2.0(4)
9.0(5)
<2.0(5)
1.7(5)
1.4(4)
1.1(4)
1.7(4)
2.0(5)
2.0(4)
7.3(5)
2.1(5)
4.6(5)
(b)
7.6(5)
...
2.0(5)
<7(6)
4.1(4)
1.9(3)
2.6(3)
6.3(5)
2(5)
1.5(4)
2.0(4)
1.7(4)
8.3(5)
<1.1(5)
2.3(5)
5.3(5)
1.1(4)
1.9(6)
...
1.3(4)
7.5(5)
7.3(5)
3.5(5)
(c)
IC 418
Orion
NGC 3576
<1.1(4)
...
...
...
4.4(4)
1.5(3)
2.3(3)
5(5)
...
1.8(4)
...
2.3(4)
...
...
...
1.2(4)
9(5)
1.4(4)
...
...
7(5)
8(5)
1.0(4)
2.8(5)
...
<4.9(5)
<5.3(5)
3.2(4)
3.5(5)
3.5(5)
<1.4(5)
<2.1(5)
<8.4(5)
...
2.0(5)
<2.3(5)
1.7(5)
<2.3(5)
1.3(4)
1.9(5)
1.6(5)
<2.8(5)
<9.3(6)
<1.8(5)
<1.5(5)
2.4(5)
...
...
...
...
7.0(5)
...
7.8(5)
...
...
...
...
...
5.6(6)
...
...
<7.8(5)
<7.8(5)
...
...
...
...
...
...
...
...
...
...
<3.0(5)
...
<5.0(5)
...
...
...
...
...
...
...
...
...
<5.0(5)
<2.0(5)
...
...
...
...
...
Notes.—In units of I(H) ¼ 1; certain identifications shown with an asterisk. NGC 7027: (a) present measurement; (b) Péquignot & Baluteau (1994); (c) Zhang et al.
(2005); Orion: Baldwin et al. (2000); NGC 3576: Garcı́a-Rojas et al. (2004).
supplemental Z > 36 ions into EMILI. This leads to an error
of up to 10 8 in the transition wavelength. Thus, the identification cannot be confirmed under the present degree of energy
level uncertainty. Also missing are the Ba ii transitions 6s 2S1/2 Y
o
o
k4554.03 and 5d 2D3/2 Y6p 2P3/2;1/2
kk5853.67, 6496.90.
6p 2P3/2
As discussed previously, all would be weaker in our spectra than
the potentially observed k6141.71 under the assumption of a
moderate optical depth in the resonance transition k4554.03 (and
o
k4934.08, not observed by PB94).
6s 2S1/2 Y6p 2P1/2
Zhang et al. (2005) have claimed the identification of five
additional transitions in their NGC 7027 spectra belonging to
Z > 30 elements that were not observed in our spectra. Two additional permitted Ba ii lines, Ba ii k4554.03 and Ba ii k4934.08,
are detected, both weaker than the k6141.71 line as would be expected for substantial optical depths of the resonance lines. They
also identify auroral [Br iii] k7385.2 and [Rb v] k5080.2, although the nebular [Br iii] k6131.0 and [ Rb v] kk4742.4, 5363.6
lines, which might be expected to be stronger than the auroral
lines, are not identified. Zhang et al. (2005) also identify [Sr vi]
k4249.2, which might be better attributable to [ Fe ii] k4249.08. It
should be noted that the present spectrum of NGC 7027 and that
of Zhang et al. (2005) were obtained at different spectral resolutions, signal-to-noise ratios, and locations in the PN, resulting
in some differences in the lines detected and identified in each.
6.5. Z > 30 Line Identifications in H ii Regions
In comparing the spectra of the PNe with H ii regions, we note
that no postYFe peak lines were identified by Garcı́a-Rojas et al.
(2004) in their deep VLT UVES echelle spectrum of the H ii region NGC 3576. In Table 11 are included estimates of the upper
limits to undetected Kr and Xe line intensities for this H ii region
from the faintest detectable lines that were identified at neighboring wavelengths or from artificial lines inserted at their wavelengths
that meet minimal detection S/N statistics. In the spectrum of
the Orion Nebula, Baldwin et al. (2000) reported weak (I 2 ;
105 H ) unidentified features at 5867.8 and 6826.9 8 that were
originally unidentified but can now be positively identified as
o
o
Y 2 D3/2
k5867.74 and [Kr iii] 3P2 Y 1D2 k6826.70,
[ Kr iv] 4 S3/2
respectively. However, there are no other close coincidences between any lines listed in Tables 7Y10, except for k5363.34, which
o
o
Y 2 D3/2
k5363.60, an unwould correspond with [Rb v] 4 S3/2
likely identification given the unrealistically high degree of ionization (40 eV ) for an H ii region. The observed intensities and
upper limits are also included in Table 11.
7. Z > 30 ABUNDANCES
The line intensities given in Table 11 were used to compute
abundances for Ba, Kr, Xe, and Br ions. The atomic data listed in
Table 3 were formatted for inclusion in five-level atom models
for abundance analysis with the IRAF NEBULAR task abundance, or for the cases of the fine-structure lines [Xe vi] k6408.9
and [Ba iv] k5696.6, using a two-level atomic solution code. For
the [Br iii] and [Br iv] lines the relevant collision strengths have
not yet been calculated. However, since these ions are isoelectronic with [Kr iv] and [Kr v], and because collision strengths for
the same levels along an isoelectronic sequence tend to vary with
effective nuclear charge (Seaton 1958), with some exceptions,
the collision strengths of [ Br iii] and [Br iv] were assumed to be
25% smaller than those for Kr. Because collision strengths were
not calculated by Schöning & Butler (1998) for [Xe v] transitions, the [Kr v] collision strengths for the same transitions were
utilized in the Xe+4 analysis. Collision strengths for excitation to
1288
SHARPEE ET AL.
Vol. 659
TABLE 12
s-Process Ionic/ Elemental Abundances
PNe
X+i/ H + Transition
+2
+
Br / H k6131.0............
Br+2/ H + k6556.4............
Adopted..........................
Br+3/ H + k7368.1............
Br+3/ H + k9450.5............
Br+3/ H + adopted ............
Kr+2/ H + k6826.70 .........
Kr+3/ H + k5346.02 .........
Kr+3/ H + k5867.74 .........
Kr+3/ H + k6107.8 ...........
Kr+3/ H + k6798.4 ...........
Kr+3/ H + adopted............
Kr+4/ H + k6256.06 .........
Kr+4/ H + k8243.39 .........
Kr+4/ H + adopted............
Xe+2/ H + k5846.77 .........
Xe+3/ H + k5709.2 ...........
Xe+3/ H + k7535.4 ...........
Xe+3/ H + adopted ...........
Xe+4/ H + k7076.8 ...........
Xe+5/ H + k6408.9 ...........
Ba+/ H + k6141.71...........
Ba+3/ H + k5696.6 ...........
Ar+2/ H + ..........................
Ar+2/ H + ..........................
Ar+4/ H + ..........................
Br/Ar ..............................
Kr/Ar ..............................
Xe/Ar..............................
[ Br/Ar] ...........................
[ Kr/Ar] ...........................
[ Xe/Ar]...........................
NGC 2440
<1.77
2:13þ0:19
0:28
2:13þ0:19
0:28
<2.21
...
<2.21
...
2:96þ0:09
0:10
2:96þ0:09
0:10
2:8þ0:2
0:3
3:6þ0:3
0:7
2.94 0.07
1:7þ0:5
1:7
...
1:7þ0:5
1:7
...
1:53þ0:12
0:15
1:46þ0:12
0:15
1:49þ0:09
0:10
...
1:17þ0:16
0:19
2:73þ0:14
0:16
<1.66
6.13 0.04
5.70 0.03
þ0:06
5:240:08
...
3.33 0.11
4:64þ0:12
0:14
...
0.07 0.11
0:37þ0:12
0:14
H ii Regions
IC 2501
IC 4191
þ0:16
2:210:19
þ0:18
2:700:22
þ0:14
2:270:15
þ0:14
2:180:17
þ0:16
2:440:21
þ0:11
2:240:13
<2.89
...
<2.89
2:4þ0:3
0:5
þ0:12
2:750:15
þ0:12
2:690:15
<3.44
...
þ0:09
2:720:10
2:1þ0:6
2:1
...
2:1þ0:6
2:1
1:3þ0:2
0:3
þ0:18
1:760:28
þ0:12
1:940:15
þ0:11
1:870:13
<3.01
<0.68
<1.85
<1.15
6.27 0.07
4.75 0.03
...
...
3.3 0.3
4:31þ0:15
0:19
...
0.1 0.3
0:04þ0:15
0:19
<2.74
...
<2.74
þ0:13
1:890:17
3:28 0:10
2.99 0.10
...
<4.03
3.07 0.07
...
...
...
...
þ0:13
1:970:15
þ0:12
2:250:15
þ0:10
2:050:11
þ0:17
1:820:27
þ0:3
0:30:4
::
2:5:0:2
<1.13
þ0:05
5:700:06
þ0:03
5:740:04
þ0:09
4:570:12
...
2:94þ0:09
0:10
3:78þ0:13
0:17
...
þ0:09
0:320:10
þ0:13
0:490:17
NGC 7027
IC 418
Orion
NGC 3576
2:52þ0:09
0:10
3:14þ0:10
0:12
2:57þ0:08
0:09
2:53þ0:09
0:11
2:74þ0:11
0:14
2:58þ0:08
0:09
3:39þ0:07
0:08
3:78þ0:02
0:03
2:41þ0:17
0:22
...
...
...
...
...
...
3.01
...
2.83
...
...
2.83
...
...
...
<2.47
<3.02
...
<3.02
...
...
...
...
6.42
4.37
...
...
3.19
<3.29
...
+0.07
<0.98
...
...
...
...
...
...
<2.47
...
<2.60
...
...
<2.60
...
...
...
...
<2.80
<2.15
<2.15
...
...
...
...
6.34 0.04a
4.20 0.07a
...
...
<3.50
...
...
<0.24
...
3.77 0.02
3:99þ0:05
0:06
3:8þ0:3
0:6
3.79 0.02
3.00
3.10
3:05þ0:05
0:06
2:2þ0:2
0:4
2.58 0.04
2:63 0:06
2.59 0.03
2:38þ0:08
0:09
2.28
::
2:41:0:05
1:7þ0:3
1:2
6.09 0.02
5.77 0.02
5.60
3:22þ0:16
0:22
2.36 0.04
3:35þ0:08
0:09
0:73þ0:16
0:22
0.90 0.04
0:92þ0:08
0:09
<3.07
2:41þ0:17
0:22
<3.00
<3.02
<3.00
3:53þ0:13
0:16
2:60þ0:17
0:18
2:40þ0:16
0:18
<4.07
<4.65
2.46 0.12
<3.15
...
<3.15
2:53þ0:13
0:16
2:32þ0:16
0:18
2:02þ0:16
0:18
2.10 0.12
<2.83
<1.24
<1.97
<1.90
6:01þ0:06
0:07
2:96þ0:12
0:06
...
...
þ0:14
2:440:18
þ0:14
3:340:17
...
0:82þ0:14
0:18
0:93þ0:14
0:17
Note.—In units of 12 þ log (N/ H).
a
Ar abundances using t 2 ¼ 0:00.
the 3P2 parent level of the tentatively observed [Xe v] 5p2 3P0 Y
3
P2 k7076.8 line in other np2 ions such as Ne v ( Lennon & Burke
1994), Ar v (Galavis et al. 1995), and Kr v (Schöning 1997)
depart by 15% at most from their Kr v values at 10,000 K. The
temperatures and densities used for these analyses corresponded
to those from diagnostics with the closest ionization potential. The [Ar iii] (27.6 eV ) temperature and [Cl iii] (23.8 eV ) density diagnostic values were used for [Br iii] (21.8 eV ), [Kr iii]
(24.4 eV ), [ Xe iii] (21.1 eV), and [Ba iv] (20.0 eV ); [O iii]
(35.1 eV) temperature and [Ar iv] (40.1 eV ) density for [ Br iv]
(36.0 eV ), [Kr iv] (37.0 eV ), and [Xe iv] (32.1 eV ); [ Ne iii]
(41.0 eV ) temperature and [Ar iv] (40.1 eV) density for [Xe iv]
(46.0 eV); and [Ar v] (59.8 eV) temperature and [K v] (60.9 eV )
density for [Kr v (52.5 eV )] and [Xe vi] (57.0 eV ). The [O i]
temperature and [N i] density were used for Ba ii assuming collision excitation of Ba+ as the source of Ba ii k6141.71. Averaged
diagnostic values were used when one of the diagnostics was
unavailable for a particular ion. Uncertainties were computed in
the same manner as the lighter elements, through permutation of
intensity measurement and diagnostic value errors, where available, and selection of extrema values. For lines that were corrected for blending, 25% of the value of the correction was added
to the measurement uncertainty.
Table 12 presents the results of the abundance determinations
for individual lines of Ba, Br, Kr, and Xe ions. To compute over-
all elemental abundances for the latter three ions, argon has been
selected as a benchmark element in addition to hydrogen. This
was done because the ionization potentials of the noble gases Ar,
Kr, and Xe, as well as Br, are very similar for their first three
stages of ionization and therefore ionization corrections should
not be large when making abundance comparisons among these
elements. In addition, the noble gases are almost completely
nonreactive and have very low condensation temperatures, so
corrections for gas-phase abundances depleted by grain formation are insignificant for these elements. Only ionic abundances
relative to hydrogen were computed for Ba.
We convert from ionic to elemental abundances by making use of the similarity in ionization potentials of the noble gases, so that Kr/Ar ¼ (Krþ2 þ Krþ3 )/(Arþ2 þ Arþ3 ) and
(Krþ2 þ Krþ3 þ Krþ4 )/(Arþ2 þ Arþ3 þ Arþ4 ) and Xe/Ar ¼
(Xeþ2 þ Xeþ3 )/(Arþ2 þ Arþ3 ) and (Xeþ2 þ Xeþ3 þ Xeþ4 þ
Xeþ5 )/(Arþ2 þ Arþ3 þ Arþ4 ) for H ii regions and PNe judged
to be of low excitation ( IC 2501 and IC 418) and for highexcitation PNe ( NGC 2440, IC 4191, NGC 7027), respectively.
The Kr+4/ H+ abundance was included in the IC 2501 Kr /Ar
ratio determination. For NGC 7027, the Br/Ar ratio was calculated the same way as the Xe/Ar ratio, with corrections for
unobservable Br+4 and Br+5 made assuming Brþ4 ; Brþ5 /Br ¼
Xeþ4 ; Xeþ5 /Xe, appropriate because of their very similar ionization potentials.
No. 2, 2007
NEBULAR s-PROCESS ABUNDANCES
The resulting abundances are listed in the bottom rows of
Table 12 relative to their solar values from Lodders (2003). Three
particular results are noteworthy. First, three out of five PNe show
significant enhancements in Kr and Xe relative to solar values,
while two others show similar, solar-like values. It is also interesting to note that IC 418 and NGC 7027, both considered young
PNe, have the largest overabundances, although they are of
greatly different ionization classes. Meanwhile, the H ii regions
show only solar Kr abundances, indicative of unprocessed interstellar medium (ISM ) gas. Secondly, the Kr and Xe abundances
in all PNe all show enhancements of similar magnitude, as was
seen in NGC 7027 by PB94. Finally, for NGC 7027, the Br abundance also shows a level of enhancement similar to that of Kr and
Xe. Uncertainties remain regarding the adaptability of Kr collision
strengths, modified as was done here, for the Br abundance calculations. However, to reduce the Br abundance to solar, the collision strengths for Br+2 and Br+3 would have to both be 4 times
greater than their counterparts in Kr, and this is much larger than
the difference seen for analogous transitions for other p 2 and p3
ion pairs (N and O, Cl and Ar). In any event, the prevalence of
probable Br line identifications suggests, independent of the actual atomic data, that significant Br does exist in most of the PNe
comprising our sample.
It is of interest to compare the s-process abundances we derive
for PNe with those obtained for evolved stars from analyses of
their absorption spectra. With the exception of Rb, Ba, and Pb,
the elements observed in emission in PNe are different from those
normally observed in the spectra of late-type stars, so a direct comparison is not feasible. In fact, lines of the same s-process elements
are not necessarily even observed in similar-type stars. For this
reason Luck & Bond (1991) defined two parameters, [ls] and
[hs], that represent the mean abundance of elements associated
with Sr and Ba, respectively, in what are called the ‘‘light’’ and
‘‘heavy’’s-process peaks. They define the abundance indices [ ls]
and [hs] for an object as the mean logarithmic abundances relative to iron of (Y, Zr, and Sr) and ( Ba, Nd, La, and Sm), respectively, compared to their solar mean abundances. All of these
elements are produced by the s-process, although several of them
have predominantly r-process contributions for solar-type stars
(Arlandini et al. 1999).
Of the postYFe peak elements that are observed in PNe, Kr
belongs to the light s-process peak near Sr, and Xe is near Ba in
the heavy s-process peak. Both are produced by the s-process,
although Xe is predominantly an r-process element for stars of
solar metallicity. Using models to determine the appropriate correction factors, Kr and Xe can be incorporated into the [ls] and
[hs] indices. However, in making comparisons of nebular abundances with those derived from stellar spectra, Fe should not be
used as the fiducial abundance for nebulae because its consistently strong depletion from the gas phase due to grain formation
causes its true abundance in nebulae to be indeterminate (Shields
1975; Perinotto et al. 1999). Argon is a good surrogate for Fe because it does not suffer depletion in nebulae, and it is well ob-
1289
served in nebulae in multiple ionization stages and the relevant
excitation cross sections are known.
The extent to which s-process nucleosynthesis occurs in
stars is determined by the total neutron exposure after the third
dredge-up phase. Calculations show that neutron exposure is the
primary factor that determines the ratio of the enhancements of
the heavy to light s-process elements, i.e., [hs/ ls], the ‘‘neutron
exposureYrelated parameter.’’ Low neutron exposures result in
element production confined to a crowded region around the Sr
peak, producing ½hs/ ls < 0, whereas higher neutron exposures
significantly populate the Ba peak, leading to ½hs/ ls > 0 (Busso
et al. 1995, 1999). The fact that ½Xe/ Kr 1 for PNe, in spite of
Xe not being produced by the s-process as much as Kr, is strongly
suggestive that the progenitor AGB stars of PNe experience significant neutron exposure. Of equal significance, the fact that forbidden lines of postYFe peak elements not observed in stars are
detectable in nebulae causes H ii regions and PNe to be of potentially great value in studying the nucleosynthesis of these elements.
8. SUMMARY
In summary, very high S/N spectra of PNe do reveal lines
from elements that are enhanced by the s-process, as was also
found by other investigators. Because many postYFe peak elements are refractory, their gas-phase abundances are not reliable
indicators of the nucleosynthetic processes that have occurred in
the progenitor stars. Fortunately, the noble gases Ar, Kr, and Xe
are largely unaffected by molecular and grain formation, and the
latter two elements are situated in the light and heavy s-process
peaks, respectively. They also have easily excitable and observable
lines in multiple ionization stages for which atomic cross sections
are now available and whose analyses should be straightforward.
We find for a sample of five PNe that Kr and Xe abundances
are enhanced over solar values by up to an order of magnitude,
from both an analysis of their intensities relative to those of the
fiducial element Ar and also a relative comparison of their line
strengths in PNe versus H ii regions such as the Orion Nebula,
where the lines are very weak in gas that is representative of the
ISM composition. The similar enhancements of Kr and Xe in
PNe are suggestive of large neutron exposures in the progenitor
central stars. Further spectroscopy of PNe should reveal additional postYFe peak emission lines whose analyses will contribute to a more complete picture of postYmain-sequence stellar
evolution, while deep spectra of H ii regions should lead to improved values of the ISM abundances of these elements.
The work of Y. Z. and X. W. L. was partially supported by
Chinese NSFC grant 10325312, and Y. Z. gratefully acknowledges the award of an Institute Fellowship from STScI, where his
work was carried out. E. P., K. C., and J. A. B. gratefully acknowledge support for this work from NSF grant AST 03-05833
and HST grant GO09736.02-A.
REFERENCES
Arlandini, C., Kappeler, F., Wisshak, K., Gallino, R., Lugaro, M., Busso, M., &
Biémont,
——— E., & Hansen, J. E. 1986b, Phys. Scr., 34, 116
Straniero, O. 1999, ApJ, 525, 886
Biémont, E., Hansen, J. E., Quinet, P., & Zeippen, C. J. 1995, A&AS, 111, 333
Baldwin, J. A., Verner, E. M., Verner, D. A., Ferland, G. J., Martin, P. G.,
Busso, M., Gallino, R., & Wasserburg, G. J. 1999, ARA&A, 37, 239
Korista, K. T., & Rubin, R. H. 2000, ApJS, 129, 229
Busso, M., Lambert, D. L., Beglio, L., Gallino, R., Raiteri, C. M., & Smith, V. V.
Baluteau, J.-P., Zavagno, A., Morisset, C., & Péquignot, D. 1995, A&A, 303, 175
1995, ApJ, 446, 775
Barge, T., Wahlgren, G. M., Johansson, S. G., Leckrone, D. S., & Proffitt, C. R.
Cahn, J. H., Kaler, J. B., & Stanghellini, L. 1992, A&AS, 94, 399
1998, ApJ, 496, 1051
Churilov, S. S., & Joshi, Y. N. 2000, Phys. Scr., 62, 358
Benjamin, R. A., Skillman, E. D., & Smits, D. P. 1999, ApJ, 514, 307
Churilov, S. S., Joshi, Y. N., & Gayasov, R. 2002, J. Opt. Soc. Am. B, 18, 113
Biémont, E., Cowan, R. D., & Hansen, J. E. 1988, Phys. Scr., 37, 850
Cosby, P. C., Sharpee, B. D., Slanger, T. G., Huestis, D. L., & Hanuschik, R. W.
Biémont, E., & Hansen, J. E. 1986a, Phys. Scr., 33, 117
2006, J. Geophys. Res., 111, A12307
1290
SHARPEE ET AL.
Cosby, P. C., & Slanger, T. G. 2007, Canadian J. Phys., in press
Davey, A. R., Storey, P. J., & Kisielius, R. 2000, A&AS, 142, 85
Dinerstein, H. L. 2001, ApJ, 550, L223
Esteban, C., Peimbert, M., Garcı́a-Rojas, J., Ruiz, M. T., Peimbert, A., &
Rodrı́guez, M. 2004, MNRAS, 355, 229
Froese Fischer, C., & Tachiev, G. 2004, At. Data Nucl. Data Tables, 87, 1
Galavis, M. E., Mendoza, C., & Zeippen, C. J. 1995, A&AS, 111, 347
Gallard, M., Raineri, M., Reyna Almandos, J. G., Sobral, H., & Callegari, F.
1999, J. Quant. Spectrosc. Radiat. Transfer, 61, 319
Garcı́a-Rojas, J., Esteban, C., Peimbert, M., Rodrı́guez, M., Ruiz, M. T., &
Peimbert, A. 2004, ApJS, 153, 501
Hamuy, M., Suntzeff, N. B., Heathcote, S. R., Walker, A. R., Gigoux, P., &
Phillips, M. M. 1994, PASP, 106, 566
Hansen, J. E., & Persson, W. 1974, J. Opt. Soc. Am., 64, 696
———. 1976, Phys. Scr., 13, 166
Hanuschik, R. W. 2003, A&A, 407, 1157
Hobbs, L. M., Thorburn, J. A., Oka, T., Barentine, J., Snow, T. P., & York, D. G.
2004, ApJ, 615, 947
Howarth, I. D. 1983, MNRAS, 203, 301
Hyung, S., Aller, L. H., Feibelman, W. A., & Lee, S.-J. 2001, ApJ, 563, 889
Hyung, S., Keyes, C. D., & Aller, L. H. 1995, MNRAS, 272, 49
Joshi, Y. N., Tauheed, A., & Davison, I. G. 1992, Canadian J. Phys., 70, 740
Karlsson, H., & Litzén, U. 1999, Phys. Scr., 60, 321
Kaufman, V., Sugar, J., & Joshi, Y. N. 1988, J. Opt. Soc. Am. B, 5, 619
Kingdon, J., & Ferland, G. J. 1995, ApJ, 442, 714
Kisielius, R., & Storey, P. J. 2002, A&A, 387, 1135
Kisielius, R., Storey, P. J., Davey, A. R., & Neale, L. T. 1998, A&AS, 133, 257
Klose, J. Z., Fuhr, J. R., & Weise, W. L. 2002, J. Phys. Chem. Ref. Data, 31, 217
Lennon, D. J., & Burke, V. M. 1994, A&AS, 103, 273
Liu, X.-W. 2006, in Planetary Nebulae beyond the Milky Way, ed. J. Walsh, L.
Stanghellini, & N. Douglas ( Berlin: Springer), 169
Lodders, K. 2003, ApJ, 591, 1220
Luck, R. E., & Bond, H. E. 1991, ApJS, 77, 515
Mendoza, C. 1983, in IAU Symp. 103, Planetary Nebulae, ed. D. R. Flower
( Dordrecht: Reidel), 143
Moore, C. E. 1952, Atomic Energy Levels, Vol. II, NBS Circ. 467 ( Washington,
DC: NBS)
———. 1958, Atomic Energy Levels, Vol. III, NBS Circ. 467 ( Washington,
DC: NBS)
Nussbaumer, H., & Storey, P. J. 1984, A&AS, 56, 293
Osterbrock, D. E., & Ferland, G. J. 2006, Astrophysics of Gaseous Nebulae and
Active Galactic Nuclei (2nd ed.; Sausalito: Univ. Science Books)
Osterbrock, D. E., Fulbright, J. P., Martel, A. R., Keane, M. J., & Trager, S. C.
1996, PASP, 108, 277
Peimbert, M., Peimbert, A., Ruiz, M. T., & Esteban, C. 2004, ApJS, 150, 431
Péquignot, D., & Baluteau, J. P. 1994, A&A, 283, 593 ( PB94)
Péquignot, D., Petitjean, P., & Boisson, C. 1991, A&A, 251, 680
Perinotto, M., Bencini, C. G., Pasquali, A., Manchado, A., Rodrı́guez Espinosa,
J. M., & Stanga, R. 1999, A&A, 347, 967
Persson, W., & Petterson, S.-G. 1984, Phys. Scr., 29, 308
Persson, W., & Reader, J. 1986, J. Opt. Soc. Am. B, 3, 959
Persson, W., & Wahlstrom, C.-G. 1985, Phys. Scr., 31, 487
Persson, W., Wahlstrom, W.-G., Bertuccelli, G., DiRocco, H. O., Reyna
Almandos, J. G., & Gallardo, M. 1988, Phys. Scr., 38, 347
Reader, J. 1983, J. Opt. Soc. Am., 73, 349
Reader, J., & Acquista, N. 1976, J. Opt. Soc. Am., 66, 896
Reader, J., & Epstein, G. L. 1972, J. Opt. Soc. Am., 62, 619
Robertson-Tessi, M., & Garnett, D. R. 2005, ApJS, 157, 371
Rodrı́guez, M. 2003, ApJ, 590, 296
Safronova, U. I., Safronova, M. S., & Johnson, W. R. 2005, Phys. Rev. A, 71,
052506
Sansonetti, C. J., Reader, J., Tauheed, A., & Joshi, Y. N. 1993, J. Opt. Soc. Am. B,
10, 7
Schöning, T. 1997, A&AS, 122, 277
Schöning, T., & Butler, K. 1998, A&AS, 128, 581
Seaton, M. J. 1958, Rev. Mod. Phys., 30, 979
Sharpee, B., Baldwin, J. A., & Williams, R. 2004, ApJ, 615, 323
Sharpee, B., Williams, R., Baldwin, J. A., & van Hoof, P. A. M. 2003, ApJS,
149, 157
Shaw, R. A., & Dufour, R. J. 1995, PASP, 107, 896
Shields, G. A. 1975, ApJ, 195, 475
Smits, D. P. 1991, MNRAS, 251, 316
———. 1996, MNRAS, 278, 683
Sterling, N. C., & Dinerstein, H. L. 2003, Rev. Mex. AA, 18, 133
———. 2004, in ASP Conf. Ser. 313, Asymmetric Planetary Nebulae III, ed.
M. Meixner et al. (San Francisco: ASP), 410
Sterling, N. C., Dinerstein, H. L., & Bowers, C. W. 2002, ApJ, 578, L55
Storey, P. J. 1994, A&A, 282, 999
Storey, P. J., & Hummer, D. G. 1995, MNRAS, 272, 41
Tauheed, A., & Joshi, Y. N. 1992, Phys. Scr., 46, 403
———. 1993a, Phys. Rev. A, 47, 3092
———. 1993b, Phys. Scr., 47, 550
Tauheed, A., Joshi, Y. N., & Kaufman, V. 1991, Phys. Scr., 44, 579
Tauheed, A., Joshi, Y. N., & Pinnington, E. H. 1993, Phys. Scr., 47, 555
Wilson, N. J., & Bell, K. L. 2002, MNRAS, 331, 389
Zethson, T., Gull, T. R., Hartman, H., Johansson, S., Davidson, K., & Ishibashi, K.
2001, AJ, 122, 322
Zhang, Y., & Liu, X.-W. 2002, MNRAS, 337, 499
Zhang, Y., Liu, X.-W., Luo, S.-G., Péquignot, D., & Barlow, M. J. 2005, A&A,
442, 249
Zhang, Y., Liu, X.-W., Wesson, R., Storey, P. J., Liu, Y., & Danziger, I. J. 2004,
MNRAS, 351, 935